second european union minorities and …...3 foreword the results of fra’s second large-scale...
Post on 13-Mar-2020
5 Views
Preview:
TRANSCRIPT
EU-MIDIS II
Second European Union Minorities and Discrimination Survey
Main results
Europe Direct is a service to help you find answers to your questions about the European Union
Freephone number (*):00 800 6 7 8 9 10 11
(*) The information given is free, as are most calls (though some operators, phone boxes or hotels may charge you).
Photo (cover & inside): © FRA icons; FRA (Helsinki and Pavlikeni); AdobeStock (Arkady Chubykin; J. Pchret; Jasmin Merdan; goodluz; Daniel Ernst; Burlingham; Andrey Arkusha; ajr images; blvdone; Rachel Kolokoff-Hopper; Alex Shevarev; Fergus Coyle; Eugene Sergeev)
More information on the European Union is available on the internet (http://europa.eu).
Luxembourg: Publications Office of the European Union, 2017
FRA – print: ISBN 978-92-9491-761-4 doi:10.2811/902610 TK-AN-17-003-EN-CFRA – web: ISBN 978-92-9491-762-1 doi:10.2811/268615 TK-AN-17-003-EN-N
© European Union Agency for Fundamental Rights, 2017
Reproduction is authorised provided the source is acknowledged. For any use or reproduction of photos contained herein, permission must be sought directly from the copyright holder.
Printed by Bietlot in Belgium
Printed on elemental chlorine-free bleached paper (ECF)
Second European Union Minorities and Discrimination Survey
Main results
3
ForewordThe results of FRA’s second large-scale European Union-wide survey on migrants and minorities – EU-MIDIS II – are striking and frustratingly persistent. Seventeen years after adoption of EU laws that forbid discrimination, immigrants, descendants of immigrants and minority ethnic groups continue to face widespread discrimination across the EU and in all areas of life – most often when seeking employment. Almost a third of all respondents who looked for a job encountered discrimination in the five years before the survey because of their ethnic or immigrant background.
For many, discrimination is a recurring experience. Hate-motivated harassment too remains a scourge, with one in four experiencing such treatment in the year before the survey. While respondents believe their ethnic or immi-grant background is the main reason for facing discrimination, they identify their names, skin colour and religion as additional triggers.
Not surprisingly, experiences with discrimination and hate-motivated harassment and violence chip away at people’s trust in public institutions and undermine feelings of attachment to their country of residence. This impedes social integration.
Refining integration measures requires relevant and comparable data. Through EU-MIDIS II, FRA collected unique comparable data, not available from other sources, based on a robust sample of over 25,500 randomly selected respondents with different ethnic minority and immigrant background in all 28 EU Member States. It follows up and expands on FRA’s first major EU-wide survey on minorities’ and migrants’ experiences – EU-MIDIS I – conducted in 2008. The survey focuses on discrimination in different settings, police stops, criminal victimisation, rights awareness and societal participation.
The findings are a wake-up call for EU and national policymakers. We hope the findings and recommendations will inspire meaningful measures to ensure the respect of fundamental rights and full inclusion of everyone living in the EU.
Michael O’FlahertyDirector
4
Country and target groups codesCountry
codeEU Member
StateCountry target
group codeTarget group
AT Austria AT – TUR Immigrants and descendants of immigrants from Turkey AT – SSAFR Immigrants and descendants of immigrants from Sub-Saharan Africa
BE Belgium BE – TUR Immigrants and descendants of immigrants from TurkeyBE – NOAFR Immigrants and descendants of immigrants from North Africa
BG Bulgaria BG – Roma RomaCY Cyprus CY – ASIA Immigrants and descendants of immigrants from AsiaCZ Czech Republic CZ – Roma RomaDE Germany DE – TUR Immigrants and descendants of immigrants from Turkey
DE – SSAFR Immigrants and descendants of immigrants from Sub-Saharan AfricaDK Denmark DK – TUR Immigrants and descendants of immigrants from Turkey
DK – SSAFR Immigrants and descendants of immigrants from Sub-Saharan AfricaEE Estonia EE – RUSMIN Russian minorityEL Greece EL – Roma Roma
EL – SASIA Immigrants and descendants of immigrants from South AsiaES Spain ES – Roma Roma
ES – NOAFR Immigrants and descendants of immigrants from North AfricaFI Finland FI – SSAFR Immigrants and descendants of immigrants from Sub-Saharan AfricaFR France FR – NOAFR Immigrants and descendants of immigrants from North Africa
FR – SSAFR Immigrants and descendants of immigrants from Sub-Saharan AfricaHR Croatia HR – Roma RomaHU Hungary HU – Roma RomaIE Ireland IE – SSAFR Immigrants and descendants of immigrants from Sub-Saharan AfricaIT Italy IT – SASIA Immigrants and descendants of immigrants from South Asia
IT – NOAFR Immigrants and descendants of immigrants from North AfricaIT – SSAFR Immigrants and descendants of immigrants from Sub-Saharan Africa
LT Lithuania LT – RUSMIN Russian minorityLU Luxembourg LU – SSAFR Immigrants and descendants of immigrants from Sub-Saharan AfricaLV Latvia LV – RUSMIN Russian minorityMT Malta MT – SSAFR Immigrants and descendants of immigrants from Sub-Saharan AfricaNL Netherlands NL – TUR Immigrants and descendants of immigrants from Turkey
NL – NOAFR Immigrants and descendants of immigrants from North AfricaPL Poland PL – RIMGR Recent immigrants PT Portugal PT – Roma Roma
PT – SSAFR Immigrants and descendants of immigrants from Sub-Saharan AfricaRO Romania RO – Roma RomaSE Sweden SE – TUR Immigrants and descendants of immigrants from Turkey
SE – SSAFR Immigrants and descendants of immigrants from Sub-Saharan AfricaSI Slovenia SI – RIMGR Recent immigrants SK Slovakia SK – Roma RomaUK United Kingdom UK – SASIA Immigrants and descendants of immigrants from
South Asia (Pakistan and Bangladesh)UK – SSAFR Immigrants and descendants of immigrants from Sub-Saharan Africa
Country groupingsEU-28 Current 28 EU Member States
5
Acronyms and abbreviationsAAPOR American Association for Public Opinion ResearchCAPI Computer-assisted personal interviewing EC European CommissionECHR European Convention on Human Rights EEA European Economic AreaEFTA European Free Trade AssociationESS European Social SurveyEQLS European Quality of Life SurveyEU European UnionEU-28 European Union, all 28 EU Member StatesEU-LFS European Union Labour Force Survey (Eurostat)EU-MIDIS (I) (First) European Union Minorities and Discrimination SurveyEU-MIDIS II Second European Union Minorities and Discrimination SurveyEurofound European Foundation for the Improvement of Living and Working ConditionsEU-SILC European Union Statistics on Income and Living ConditionsFRA European Union Agency for Fundamental RightsID IdentificationILO International Labour OrganizationISCED International Standard Classification of EducationLGBT Lesbian, gay, bisexual and transgender(S)ASIA South Asia and AsiaSDGs Sustainable Development Goals (United Nations)TFEU Treaty on the Functioning of the European UnionUN United NationsUNESCO United Nations Educational, Scientific and Cultural Organization
7
FOREWORD ������������������������������������������������������������������������������������������������������������������������������������������������������������������������������ 3
COUNTRY AND TARGET GROUPS CODES ������������������������������������������������������������������������������������������������������������������������������� 4
ACRONYMS AND ABBREVIATIONS ���������������������������������������������������������������������������������������������������������������������������������������� 5
FIGURES AND TABLES ������������������������������������������������������������������������������������������������������������������������������������������������������������� 8
WHY IS THIS SURVEY NEEDED? �������������������������������������������������������������������������������������������������������������������������������������������� 10
1 KEY FINDINGS AND FRA OPINIONS ����������������������������������������������������������������������������������������������������������������������������� 131.1. Discrimination and awareness of rights ................................................................................................................ 131.2. Harassment and violence motivated by hatred ................................................................................................... 161.3. Police stops and treatment by the police .............................................................................................................. 181.4. Living together in the EU: citizenship, participation, trust and tolerance ........................................................ 18
2 WHAT DO THE RESULTS SHOW? ����������������������������������������������������������������������������������������������������������������������������������� 212.1. Discrimination and awareness of rights ................................................................................................................ 21
2.1.1. Discrimination experiences ......................................................................................................................... 222.1.2. Reporting discrimination ............................................................................................................................ 422.1.3. Awareness of support organisations, equality bodies and laws addressing discrimination .......... 50
2.2. Hate crime – harassment and violence ................................................................................................................. 562.2.1. Experiences of harassment motivated by hatred ................................................................................... 572.2.2. Experiences of physical violence motivated by hatred .......................................................................... 632.2.3. Harassment and physical violence against respondents’ family or friends – incidents
motivated by hatred ................................................................................................................................... 672.3. Police stops ................................................................................................................................................................ 68
2.3.1. Encounters with law enforcement ............................................................................................................ 692.3.2. Differences in police stops by gender and age ....................................................................................... 712.3.3. Circumstances and nature of most recent police stop .......................................................................... 742.3.4. Treatment by police during stops ............................................................................................................. 74
2.4. Living together: residence and citizenship, socio-economic situation, trust and tolerance ........................ 762.4.1. Residence status and citizenship .............................................................................................................. 792.4.2. Education and language proficiency ........................................................................................................ 892.4.3. Labour market participation ...................................................................................................................... 952.4.4. Trust in public institutions and political participation ............................................................................ 992.4.5. Sense of belonging, attachment and social distance .......................................................................... 1042.4.6. Effect of discrimination and victimisation on sense of belonging and trust in public
institutions .................................................................................................................................................. 109
ANNEX I: EU-MIDIS II METHODOLOGY ������������������������������������������������������������������������������������������������������������������������������ 114
ANNEX II: RESPONDENTS IN THE EU-MIDIS II SURVEY ���������������������������������������������������������������������������������������������������� 118
ANNEX III: AWARENESS OF EQUALITY BODIES IN THE EU-MIDIS II SURVEY ����������������������������������������������������������������� 123
REFERENCES ������������������������������������������������������������������������������������������������������������������������������������������������������������������������ 125
Contents
8
Figure 1: Grounds of discrimination experienced in four areas of daily life in 5 years before the survey (%) ............ 23Figure 2: Discrimination on different grounds in four areas of daily life in 5 years before the survey, by
age groups (%) ............................................................................................................................................................. 24Figure 3: Grounds of discrimination experienced in four areas of daily life in 5 years before the survey,
by target group (%) ..................................................................................................................................................... 25Figure 4: Grounds of discrimination experienced in four areas of daily life in 5 years before the survey,
by target group and Member States (%) ................................................................................................................. 26Figure 5: Overall discrimination based on ethnic or immigrant background in 12 months before the
survey, by survey target group and country (%) .................................................................................................... 31Figure 6: Number of discrimination experiences based on ethnic or immigrant background at work in 12
months before the survey (%) .................................................................................................................................. 33Figure 7: Number of discrimination experiences based on ethnic or immigrant background when looking
for work in 12 months before the survey (%) ......................................................................................................... 33Figure 8: Discrimination based on ethnic or immigrant background in different areas of life in 12 months
and 5 years before the survey (%) ........................................................................................................................... 34Figure 9: Discrimination based on ethnic or immigrant background in different areas of life in 12 months
before the survey, by survey target group (%) ...................................................................................................... 35Figure 10: Discrimination based on ethnic or immigrant background in ‘other public/private services’ and
in ‘education’ (self or as a parent or guardian) in 12 months before the survey, by survey target group and country (%) ................................................................................................................................................ 37
Figure 11: Discrimination based on ethnic or immigrant background when ‘looking for work’ and ‘at work’ in 12 months before the survey, by survey target group and country (%) ........................................................ 38
Figure 12: Main reason for last incident of discrimination because of skin colour, ethnic origin or religion in 5 years before the survey, in five areas (multiple response) (%) ....................................................................... 40
Figure 13: Respondents who reported or filed a complaint about last incident of discrimination, by target group and country (%) ................................................................................................................................................ 43
Figure 14: Respondents who reported or filed a complaint about last incident of discrimination, by target group and gender (%) ................................................................................................................................................. 45
Figure 15: Domains of daily life where last incident of discrimination based on ethnic or immigrant background was reported, by gender (%) .............................................................................................................. 46
Figure 16: Reports or complaints made about last incident of discrimination because of skin colour, ethnic origin or religion in all domains of daily life within SSAFR, TUR, NOAFR and ROMA target groups, all respondents, weighted and sorted by overall reporting rates (%) ............................................................... 47
Figure 17: Awareness among all respondents of organisations that offer support or advice to victims of discrimination (regardless of the grounds of discrimination), by target group (%) .......................................... 51
Figure 18: Knowledge among all respondents of at least one equality body, by country and gender (%) ................... 52Figure 19: Knowledge among all respondents of at least one equality body, by target group and country (%) .......... 54Figure 20: Awareness among all respondents of laws prohibiting discrimination based on skin colour,
ethnic origin or religion, by target group and Member States (%) ..................................................................... 55Figure 21: Prevalence of harassment due to ethnic or immigrant background in 12 months before the survey (%) .. 59Figure 22: Types of hate-motivated harassment experienced in 12 months before the survey (%) ............................. 61Figure 23: Reporting hate-motivated harassment to authorities or services – most recent incident in
5 years before the survey (%) ................................................................................................................................... 63Figure 24: Ten groups with the highest rates of hate-motivated physical violence in 12 months before
the survey, out of all groups surveyed (%) ............................................................................................................. 64Figure 25: Reasons given for not reporting hate-motivated violence to police (%) .......................................................... 67Figure 26: Prevalence of stops by police in 5 years before the survey, by EU Member State and target group (%) .... 70Figure 27: Most recent police stop being perceived as ethnic profiling among those who were stopped in
5 years before the survey, by EU Member State and target group (%) ............................................................. 72Figure 28: Most recent police stop being perceived as ethnic profiling among those who were stopped in
5 years before the survey, by gender and target group (%) ................................................................................ 73Figure 29: Countries and regions of birth of first-generation immigrants, by target group and sample size ................ 80Figure 30: Status upon arrival (left column %), current residence status (right column %) and changes of
status (thickness of lines) of immigrants ................................................................................................................ 81Figure 31: Current legal residence status among first-generation immigrants, by target group (%) .............................. 82Figure 32: First-generation respondents with secure residence status, by length of residence, target group
and country (%) ........................................................................................................................................................... 83Figure 33: National citizenship among first-generation immigrants, by gender, target group and country (%) ........... 84Figure 34: Reasons for not applying for citizenship among first-generation non-national immigrants, by
target group (%) .......................................................................................................................................................... 85
Figures and tables
9
Figures and tables
Figure 35: Status upon arrival and current residence status for first-generation immigrants from Turkey by Member State (%) .................................................................................................................................................. 87
Figure 36: Respondents aged 16–64 years who have attained upper secondary, post-secondary non-tertiary or tertiary education (ISCED 2011 levels 3–8) in any country, compared with the general population (Eurostat 2016), by target group and EU Member State (%) ............................................................ 90
Figure 37: Respondents aged 16–64 years who have completed at most lower secondary education (ISCED 2011 levels 0-2) in any country and do not continue with further education or training, by target group and age (%) ....................................................................................................................................................... 92
Figure 38: Respondents with ‘good to mother tongue’ language proficiency (in all three dimensions – speaking, reading and writing) of at least one national language in the country where interviewed, by target group and EU Member State (%) ..................................................................................... 93
Figure 39: Respondents with ‘good to mother tongue’ language proficiency in the country of residence, by level of education (ISCED) and target group (%) .................................................................................................... 95
Figure 40: Paid work rate for household members aged 20–64 years, including self-employment and occasional work or work in the past 4 weeks, compared with the Europe 2020 employment rate 2015 (Eurostat), by target group and EU Member State (%) ................................................................................ 96
Figure 41: Paid work rate for respondents aged 20–64 years, by education level and target group (%) ..................... 98Figure 42: Levels of trust in the police, by country and target group (average value on a scale from 0 to 10) ............ 101Figure 43: Levels of trust in the legal system, by country and target group (average value on a scale from
0 to 10) ......................................................................................................................................................................... 102Figure 44: Levels of trust in the legal system, by country, target group and generation ................................................ 103Figure 45: Identification with country of residence, by country, target group and generation (%) .............................. 105Figure 46: Respondents’ comfort level with other ethnic minorities and gay, lesbian or bisexual persons as
neighbours, by country and target group (%) ...................................................................................................... 107Figure 47: Respondents who have friends without an ethnic minority background (%) ................................................ 108Figure 48: Impact of experiences with discrimination, harassment and violence on trust in the legal
system (mean value on a scale from 0 to 10) ........................................................................................................ 111Figure 49: Impact of experiences with discrimination on trust in the legal system, by country and target
group (mean value on a scale from 0 to 10) .......................................................................................................... 112Figure 50: Regression analysis on the level of trust in the legal system of first-generation migrants, by
target group ................................................................................................................................................................. 113Figure 51: Number of first-generation immigrants from selected target groups, by countries covered in
EU-MIDIS II .................................................................................................................................................................... 116Figure 52: Confidence intervals of estimates for selected indicators, by country and target group (%) ....................... 117
Table 1: Most common reasons for not reporting an incident in different areas, top three per area and top six overall, sorted by overall mention (%) ....................................................................................................... 49
Table 2: Language proficiency of respondents (speaking, reading and writing) in at least one national language in the country of residence, by target group and generation (%) ..................................................... 94
Table 3: Language proficiency with regard to speaking, reading and writing the national language in their country, for Roma and Russian minority (%) ................................................................................................ 94
Table 4: Paid work rate for respondents aged 20–64 with sufficient and insufficient language proficiency (in all three dimensions – speaking, reading and writing) of at least one national language in the country where interviewed, by target group (%) ..................................................................... 99
Table 5: Feeling of attachment to country of residence and country of origin for first- and second-generation respondents (%) .................................................................................................................................... 106
Table 6: Agreement to gender equality statement “Both husband and wife should contribute to household income”, by gender, country and target group (%) .......................................................................... 110
Table 7: Main characteristics of EU-MIDIS II target groups, by country ........................................................................... 119Table 8: Most important countries of origin of first-generation immigrants per country and target group ............ 120Table 9: Awareness of equality bodies in each EU Member State (%) ............................................................................. 123
Second European Union Minorities and Discrimination Survey – Main results
10
Why is this survey needed?The population of the European Union (EU) is already highly diverse and is becoming more so. Alongside established minorities – such as the Roma and national minorities – immigration to the EU has played a signifi-cant role in shaping ethnic and cultural diversity in the Union. This has triggered challenges to social cohesion and respect for human rights, prompting the gradual development of policy and legal measures in many Member States and at EU level.
More specifically, in 2000, the EU adopted legisla-tion on equal treatment irrespective of racial or eth-nic origin, as well as legislation establishing a general framework for equal treatment in employment and occupation irrespective of religion or belief, disability, age or sexual orientation. These legal measures were accompanied by a range of policy measures aiming to strengthen the provisions’ implementation and promote efforts to tackle discrimination. In addition, in 2008, the EU adopted legislation on combating certain forms and expressions of racism and xenophobia by means of criminal law.
However, effectively assessing the impact of these measures on the ground proved difficult due to the absence of relevant and comparable official data.1
The need for data: filling the gap
To make up for this gap, in 2008, the agency conducted the first-ever major EU-wide survey on minorities’ and migrants’ experiences with discrimination and criminal victimisation. This provided, for the first time, robust empirical evidence in the form of comparable statisti-cal data across EU Member States – an invaluable basis for solid assessments of the impact of legal and policy measures on the ground and for EU and Member State efforts to improve their legal and policy responses.
The second wave of this survey, conducted in 2015 and 2016, provides evidence on how the situation has developed over the past years – covering additional areas such as citizenship, residence, participation, trust and tolerance. FRA published the first results of this second wave – focusing on Roma in selected Member
1 European statistical instruments that collect data on immigrant integration, such as the EU Labour Force Survey (EU-LFS) and the EU Statistics on Income and Living Conditions (EU-SILC), do not include questions on experiences of discrimination or bias motivated criminal victimisation. They are also not designed to regularly capture immigrant or minority populations in their samples. National data sources are not comparable across the EU due to differences in the definitions they use or in their design.
States – in November 2016.2 A second report – published in September 2017 – focused on Muslim immigrants and descendants of Muslim immigrants.3 This report pre-sents the findings for all groups surveyed. It is published together with a detailed Technical Report. The data are also available online – through a data visualisation tool on FRA’s website.
Using EU-MIDIS data to bolster diverse initiatives
EU-wide comparable equality data, as collected through the EU-MIDIS surveys, can be used to monitor the situ-ation of immigrants and the outcomes of integration policies. For example, the EU has identified ‘active citi-zenship’ and ‘welcoming society’ as key areas of inte-gration. The former is one of four policy areas regarding which Member States agreed to develop so-called “Zaragoza indicators”,4 introduced in 2010 at a minis-terial conference under the Spanish presidency of the EU. The latter has been proposed as an additional area for indicator development, and would cover discrimi-nation.5 These areas are often not entirely, or only to a very limited extent, captured in national data sources.6 EU-MIDIS II data regarding these two important areas of indicator development can provide added value to existing international comparisons of integration out-comes, such as those regularly produced by the OECD in its reports on Indicators of Immigrant Integration7 and by Eurostat.8
Regularly collecting more and better disaggregated equality data can also enhance measurement of dis-crimination and inequalities both within and among countries. In addition, such statistics can serve to chal-lenge negative assumptions and stereotypes about ethnic minorities and immigrants.
United Nations Sustainable Development Goals: gauging progress
Finally, as the United Nations (UN) Office of the High Commissioner for Human Rights (OHCHR) highlights in its note on ‘A Human Rights Approach to Data Disag-gregation to Leave No One Behind’9, such data can form the basis for analysis of the progress made regarding the UN Sustainable Development Goals (SDGs) on the most vulnerable and marginalised populations.
Non-discrimination is a core UN principle, and is embed-ded in the SDGs to be achieved by 2030. The European
2 European Union Agency for Fundamental Rights (FRA) (2016).3 FRA (2017b).4 See the Commission’s webpage on migrant integration and
the Zaragoza indicators.5 See Huddleston, Thomas et al., (2013).6 Eurostat (2011).7 OECD and European Union (2015).8 Eurostat (2017).9 See the OHCHR webpage on the note.
11
Why is this survey needed?
Commission – in its Communication of 22 Novem-ber 2016 on the ‘Next steps for a sustainable European future; European action for sustainability’ – committed itself to mainstreaming the SDGs into EU policies and initiatives, and to treat the three pillars of sustain-able development – social, environmental and eco-nomic concerns – as essential guiding principles for all its policies. The Commission also committed itself to regularly reporting, as of 2017, on the EU’s progress in reaching these goals.
Data generated through the EU-MIDIS surveys could help populate the relevant indicators, particularly on Goal 10 (‘reducing inequalities’) and Goal 16 (‘peace-ful and inclusive societies’). Relevant chapters of this report refer to specific indicators linked to these and other relevant SDGs. Furthermore, all EU-MIDIS II survey findings outlined in this report are disaggre-gated by sex/gender – and so contribute to measur-ing progress on Goal 5 (‘achieve gender equality and empower all women and girls’), particularly Target 5.1 (‘end all forms of discrimination against all women and girls everywhere’).
EU-MIDIS II in a nutshell10
n Coverage – EU-MIDIS II collected information from 25,515 respondents with different ethnic minority and immigrant backgrounds across all 28 EU Mem-ber States.
n The EU-MIDIS II sample is representative for the se-lected population groups that were surveyed.11 The sample includes groups with persons belonging to ethnic or national minorities, Roma and Russians, as well as persons born outside the EU (first-gen-eration respondents) and individuals with at least one parent born outside the EU (second-generation respondents). All respondents were aged 16 years or older, and had lived in private households for at least 12 months before the survey. Persons living in institutional settings – for example, in hospitals or prisons – were not surveyed.
n Countries/regions of origin of immigrants and descendants of immigrants include Turkey, North Africa, Sub-Saharan Africa, and South Asia (in Cy-prus, Asia); in Slovenia and Poland, individuals who immigrated to the EU in the past 10 years were in-cluded regardless of country of origin.
10 For more details on the survey methodology, see Annex I.11 In Luxembourg, it was not possible to access the available
register for sampling, so FRA applied quota sampling. Results should therefore be interpreted with caution.
n Sample characteristics – respondents are, on aver-age, 40 years of age (Russian minority respondents are on average 51 years old and recent immigrants 36 years old). Women constitute 51 % of the en-tire sample, with differences across aggregate tar-get groups and countries. In Malta and Greece, the share of female immigrants interviewed was very low – at 6 % and 5 %, respectively. The respond-ents’ socio-demographic profiles vary considerably across countries of residence and countries/regions of origin. More details on the sample characteristics can be found in Table 7 in Annex II.
n Issues covered – the survey includes questions on experiences of perceived discrimination in differ-ent settings, such as in (access to) employment, education, housing, health, and when using public or private services; on experiences of police stops, criminal victimisation (including hate crime); on awareness of rights and redress mechanisms; and on societal participation and integration, including trust in public institutions and level of attachment to the country of residence. Respondents also pro-vided information about basic socio-demographic characteristics for all household members, includ-ing themselves.
n Presentation of results – this report compares ag-gregated results on ‘target groups’ – for example, Roma or immigrants and descendants of immi-grants from North Africa, etc. – where they were surveyed in more than one EU Member State. It also presents notable findings for specific groups in in-dividual Member States. Differences with regard to gender, age or immigrant generation are presented where relevant. Where results for the first/second generation are presented, respondents with Roma and Russian ethnic minority background are exclud-ed from analysis. In a few cases, results are based exclusively on first-generation respondents due to the particular composition of the target groups in that regard in some EU Member States (see Table 7 in Annex II).
n Weighting – the survey results presented in this re-port are based on weighted data to reflect the se-lection probabilities of each household and individ-ual based on the sampling design. The weights also account for the differences in the (estimated) size of the target population in each of the countries. Where possible, the sample was post-stratified to the regional distribution and population character-istics of the covered target population. For statistics produced in this report, the samples were weighted by their estimated size, which means that country and group comparisons take the estimated total size of the target groups per country into account and do not (directly) reflect the sample sizes.
Second European Union Minorities and Discrimination Survey – Main results
12
n Comparison to EU-MIDIS I and other surveys – re-sults are compared with respect to substantial dif-ferences for selected indicators. Improvements in the sampling methodology and the application of sample design weights for the analysis of 2016 data restrict direct comparability of all results (for de-tails, see the EU-MIDIS II Technical Report and more information on the EU-MIDIS II methodology in An-nex I). Comparisons to general population surveys are included, where relevant data are available.
On terminologyBias motivation
This concerns violence and offences motivated by negative, often stereotypical, views and attitudes towards a particular group of persons who share a common characteristic, such as sex, race, ethnic origin, language, religion, nationality, sexual orien-tation, gender identity or other characteristic, such as age or a physical or psycho-social impairment. In this report, bias motivation refers to incidents of harassment and crime motivated by hatred based on respondents’ religion or religious beliefs, their ethnic or immigrant background or their skin colour.
Ethnic or immigrant background
The findings presented here use, as a generic term, ‘ethnic or immigrant background’ to include results for three grounds of discrimination asked about in the survey: skin colour, ethnic origin or immigrant background, and religion or religious belief. For more details on the intersection of ‘reli-gion’ and ‘ethnic origin’ as grounds of discrimina-tion, see Section 2.1 on ‘Discrimination and aware-ness of rights’.
Roma
The Council of Europe uses ‘Roma’ as an umbrel-la term. It refers to Roma, Sinti, Kale and related groups in Europe, including Travellers and the Eastern groups (Dom and Lom), and covers the wide diversity of the groups concerned, includ-ing persons who identify themselves as Gypsies. For the purpose of the EU-MIDIS II survey, ‘Roma’ refers to autochthonous ‘Roma’ within selected EU Member States and does not focus on ‘Roma’ who have moved from one EU Member State to another.
For more information, see Council of Europe (2012), Descriptive glossary of terms relating to Roma is-sues, Strasbourg.
13
1 Key findings and FRA opinions
The results show little progress compared to eight years ago, when the first wave of this survey was conducted: the proportions of those experiencing discrimination, as well as physical violence and harassment moti-vated by hatred, and of those not aware of relevant legislation and possibilities for redress, remain at levels that raise serious concern. Overall, Roma respondents and respondents with Sub-Saharan or North-African backgrounds – and in particular second-generation respondents – experience higher rates of discrimina-tion, harassment and violence motivated by hatred.
Nonetheless, the majority of respondents feel strongly attached to the country they live in and show high levels of trust in their country’s public institutions, including its legal system and the police. However, respondents who have experienced discrimination, harassment or violence motivated by hatred show significantly lower levels of trust and feel less attached to the country in which they live. This shows that a failure to deliver effective protection from discrimination and hate crime can undermine integration and social inclusion policies, affecting the social cohesion of our societies.
The following legislative and policy measures should be kept in mind when reading the key findings and opinions:
• The comprehensive anti-discrimination legislative framework adopted by the European Union (EU).
• The EU’s Common Basic Principles for immigrant in-tegration policy adopted in 2004 and the goals set out in the June 2016 European Commission Action Plan on the integration of third-country nationals.
• The targets in the EU’s growth strategy ‘Europe 2020’.
• The Recommendations of the Council of the EU on effective Roma integration measures in the
Member States12 and the 2011 EU Framework for National Roma Integration Strategies.13
1�1� Discrimination and awareness of rights
EU-MIDIS II results show that a considerable propor-tion of respondents face high levels of discrimination because of their ethnic or immigrant background, as well as potentially related characteristics, such as skin colour and religion. Four out of 10 respondents (38 %) felt discriminated against in the five years before the survey because of their ethnic or immigrant background in one or more areas of daily life, and one in four (24 %) experienced this in the 12 months preceding the survey. In EU-MIDIS I, one in three respondents (30 %) stated that they felt discriminated against because of their ethnicity (with respect to one or more areas of life) during the equivalent period.
Among all groups surveyed – and similarly to the find-ings of EU-MIDIS I – respondents with North African background, Roma respondents and respondents with Sub-Saharan African background continue to indicate the highest levels of discrimination based on ethnic or immi-grant background. This is the case both in the five years before the survey (at 45 %, 41 %, and 39 %, respec-tively) and in the 12 months before the survey (at 31 %, 26 %, and 24 %, respectively). Furthermore, respondents pointed to their skin colour and their first or last names as grounds of discrimination in all areas of life. Roma respondents and respondents with Sub-Saharan Afri-can background mostly experience discrimination based on their physical appearance, while immigrants and descendants of immigrants from North Africa and Turkey
12 Council of the European Union (2013).13 Council of the European Union (2011).
Second European Union Minorities and Discrimination Survey – Main results
14
more often face discrimination based on their names. Similarly to EU-MIDIS I, many EU-MIDIS II respondents describe discrimination as a recurring experience – those who have felt discriminated against indicate that this happens, on average, at least 4.6 times a year.
These findings clearly show that, eight years after the first EU-MIDIS survey in 2008, discrimination continues to affect large numbers of ethnic minorities, immigrants, and children of immigrants in the EU. This is the case even though non-discrimination is one of the Union’s fundamental principles, anchored in Article 2 and Article 3 of the Treaty on the EU (TEU), Article 10 and Article 19 of the Treaty on the Functioning of the EU (TFEU), and Article 21 of the EU Charter of Fundamental Rights.
In 2000, the EU adopted specific legislation that for-bids discrimination on grounds of racial or ethnic origin (Council Directive 2000/43/EC implementing the prin-ciple of equal treatment irrespective of racial or ethnic origin, hereinafter the Racial Equality Directive) and legislation that establishes a general framework for equal treatment in employment and occupation (Council Directive 2000/78/EC establishing a general framework for equal treatment in employment and occupation, hereinafter the Employment Equality Directive). Both the Racial Equality Directive and the Employment Equal-ity Directive prohibit various forms of discrimination: direct and indirect discrimination, harassment, instruc-tion to discriminate and victimisation. Many Member States provide even wider protection against discrimi-nation under their national legislation.
The EU-MIDIS II results underscore that having adopted non-discrimination legislation is not enough. Measures to combat discrimination in all areas of life need to be effec-tive and inclusive for all, in particular for groups most at risk of abuse – such as ethnic minorities, persons of differ-ent skin colour or religion, immigrants and their children.
FRA opinion 1 EU Member States should significantly improve the effectiveness of the measures and institutional arrangements for enforcing EU and national anti-discrimination legislation. In particular, Member States should ensure that sanctions are sufficiently effective, proportionate and dissuasive. EU Member States should also raise awareness of anti-discrimination legislation and the relevant redress mechanisms, particularly among those most likely to be affected, such as members of minority ethnic groups, as required by the Racial Equality Directive, and repeatedly called for by FRA.
The results show that a considerable proportion of respondents feel discriminated against based on their religion, particularly in employment – even though Article 10 of the EU Charter of Fundamental Rights
guarantees the freedom of thought, conscience and reli-gion, including the right to manifest religion or belief, in worship, teaching and practice. Moreover, the Employ-ment Equality Directive bans discrimination on grounds of religion or belief in employment or occupation.
As was found in EU-MIDIS I, ‘ethnic origin or immigrant background’ remains the most common ground for dis-crimination in respondents’ daily lives – affecting 25 % of all EU-MIDIS II respondents during the five years pre-ceding the survey. Religion and skin colour are also sig-nificant triggers of discrimination (both experienced by 12 %). The survey asked about encountering discrimina-tion in various areas of daily life, such as when looking for work or at work, in access to housing, and when in contact with school authorities as parents or guardians.
Discrimination is experienced differently by women and men, the young and the old, and by immigrants and descendants of immigrants. For example, on average, the second generation indicates higher levels of reli-gious discrimination than the first generation of immi-grants: one out of five second-generation respondents (20 %) felt discriminated against because of their religion or religious beliefs, compared to one out of eight first-generation immigrants (12 %). This shows that characteristics such as gender, age or socialisa-tion patterns (first and second generation) also affect discrimination experiences and need to be taken into account when designing legal and policy responses.
FRA opinion 2 EU Member States should take all necessary measures to combat religious discrimination in all areas of life. They should also take into account that women with ethnic or immigrant backgrounds and children of immigrants are most vulnerable to such discrimination. In particular, Member States should take into account that discrimination on grounds of race or ethnicity may include aspects of religious discrimination.
The EU should extend the protection against discrimination on the grounds of religion or belief, disability, age and sexual orientation to areas of social protection, including social security and healthcare, education, and access to goods and services available to the public. Adopting the proposed Equal Treatment Directive will also strengthen the legal protection against the many and pervasive forms of multiple and intersectional discrimination particularly affecting women belonging to ethnic and religious minorities, such as Roma or Muslims, in many areas of life.
Similarly to EU-MIDIS I, in EU-MIDIS II, the highest five-year rate of discrimination based on ethnic or immigrant background is indicated in the area of employment (when respondents look for work or at the workplace). Specifically, 29 % of all respondents who looked for a job
15
Key findings and FRA opinions
and 22 % of those at work in the five years before the survey felt discriminated against on this basis. Mean-while, 12 % experienced this when looking for work in the year before the survey, and 9 % did so at work dur-ing this timeframe. As in EU-MIDIS I, Roma respondents and respondents with North African background seem particularly affected by discrimination in employment (in the 12 months preceding the survey: 16 % and 15 %, respectively). Discrimination at the workplace in the 12 months preceding the survey is mostly mentioned by respondents with North African and Sub-Saharan Afri-can backgrounds (14 % and 9 %, respectively).
When asked what they thought the reason was for the most recently experienced incident of discrimination on grounds of ethnic or immigrant background in employ-ment, every second respondent who experienced such discrimination mentioned skin colour or physical appear-ance (53 % at the workplace and 50 % when looking for work ). Meanwhile, 36 % believed that the main reason for being discriminated against when looking for work was their first or last name. 18 % thought that they were discriminated against when looking for work because of their accent, and 15 % thought this was the reason for encountering discrimination at work. 12 % of respondents who felt discriminated against when looking for work – mostly respondents with North African background, espe-cially those living in France, and Roma – identified the reputation of their neighbourhood or their address as the main reason for their most recent experience of discrimi-nation. This specific reason is particularly important for Roma who experienced discrimination when looking for work in Slovakia (21 %), Greece (20 %), and Spain (18 %).
The Europe 2020 strategy is the EU's agenda for growth and jobs for the current decade. It emphasises smart, sustainable and inclusive growth as a way to over-come the structural weaknesses in Europe's economy, improve its competitiveness and productivity and underpin a sustainable social market economy. Reduc-ing the number of people threatened by poverty or social exclusion by 20 million is one of the five targets of this agenda. This is of particular importance for Roma, who are overrepresented among those affected by pov-erty and social exclusion. The Europe 2020 strategy also identified better integration of migrants as contributing towards reaching its headline target to increase the employment rate of the population aged 20–64 to 75 %, which would also foster cohesive societies and help to balance the effects of ageing populations.
Given the share of immigrants and ethnic minorities among today’s working age populations in EU Mem-ber States, these findings are particularly useful when designing policies aiming to increase the labour mar-ket participation of immigrants and minorities who can make a substantial contribution to reaching Europe 2020’s employment target.
FRA opinion 3 To reach the Europe 2020 headline target of increasing the employment rate to 75 % and help balance the effects of ageing populations, EU Member States should combat discrimination based on ethnic or immigrant background and comply with equal treatment legislation. This can reduce the barriers ethnic minorities, immigrants and their offspring face both when looking for work and at work.
To combat potential unconscious bias towards persons belonging to minority ethnic groups and to ensure equal access and labour market participation, measures could include: introducing name-blind recruitment policies; monitoring discriminatory practices; raising awareness and training on unconscious bias; supporting employers and social partners in combating discrimination and obstacles to labour market participation; and providing anti-discrimination training to employers in private companies and public services.
The EU institutions should swiftly proclaim the European Pillar of Social Rights, which includes – under the principle of equal opportunities – measures to prevent, correct and compensate for disadvantages linked to certain protected grounds. These include positive action and incentives – for instance, by supporting workforce diversity practices among employers.
EU-MIDIS I revealed that only a small proportion of respondents (18 %) reported incidents of discrimination they had experienced in the 12 months preceding that survey. EU-MIDIS II results show that the situation has not improved. To the contrary: only 12 % of respondents who felt discriminated against reported the most recent incident. This shows that measures taken to date by the EU and its Member States have not improved reporting rates. As a result, incidents of discrimination remain largely unreported and therefore invisible to institutions – such as bodies for the promotion of equal treatment – that have a legal obligation to respond to discrimination complaints.
Women report such incidents more often (14 %) than men (11 %). Respondents who did report discrimination incidents mostly addressed their employer (36 %) or trade unions (13 %) with respect to work-related inci-dents. Meanwhile, 17 % of respondents reported inci-dents related to entering a night club or a bar to the police. Only 4 % of all respondents who reported a discrimination incident with respect to any area filed a complaint or reported the incident to an equality body. This is not surprising given that most respondents are not aware that such bodies exist in their country. EU-MIDIS I showed that most respondents were not aware of any organisation that offers support or advice to discrimi-nation victims. Similarly, in EU-MIDIS II, the majority of respondents (71 %) were also not aware of any such organisation, and 62 % did not even recognise the name of any equality body in their country.
Second European Union Minorities and Discrimination Survey – Main results
16
FRA opinion 4 EU Member States should ensure that equality bodies can fulfil their tasks as assigned by the Racial Equality Directive. This means that equality bodies should be provided with the necessary staff and human resources, as required by the Racial Equality Directive and called for by FRA in its Opinion on the implementation of the equality directives (FRA Opinion 1/2013). This would enable them to:
1) receive and effectively process complaints (including complaints by third parties) and assist victims of discrimination;
2) publish independent reports and recommen-dations on any issues related to discrimination;
3) collect data through independent surveys, which provides the evidence base for monitoring levels of discrimination and awareness of the existence of equality bodies among the populations they were set up to serve.
In parallel, EU Member States should implement Article 10 of the Racial Equality Directive on dissemination of information. This means that, based on the evidence provided by EU-MIDIS II, relevant awareness-raising measures should specifically target those persons and groups vulnerable to discrimination, such as those belonging to ethnic or religious minorities.
1�2� Harassment and violence motivated by hatred
The survey asked respondents about their experiences of harassment, defined as a range of actions that the respondent found ‘offensive’ or ‘threatening’, namely offensive or threatening comments in person; threats of violence in person; offensive gestures or inappro-priate staring; offensive or threatening emails or text messages (SMS); and offensive comments made about them online. One in four respondents (24 %) experi-enced hate-motivated14 harassment, and 3 % experi-enced a hate-motivated physical attack in the 12 months before the survey.
The EU has put in place legal and policy measures to tackle hate crime, the most severe expression of discrimination and a core fundamental rights abuse. In 2008, the Council adopted Framework Decision 2008/913/JHA on combating certain forms and expres-sions of racism and xenophobia by means of criminal law. In 2016, following up on the first Colloquium on Fundamental Rights on ‘Tolerance and respect: pre-venting and combating antisemitic and anti-Muslim hatred in Europe’, the European Commission set up a EU High Level Group on combating racism, xenophobia
14 Harassment or violent incidents that respondents believe happened because of their ethnic or immigrant background.
and other forms of intolerance. This group brings together Member States, civil society and community representatives, FRA, as well as international organi-sations active in this area, to step up cooperation and improve coordination between relevant actors, maxi-mizing concrete impact on preventing and combating hate crime and hate speech on the ground.
EU-MIDIS II results show that Roma experienced the highest rate (30 %) of hate-motivated harassment in the 12 months before the survey. Respondents with North African background (29 %) indicate simi-lar rates. Comparing the results of EU-MIDIS I (2008) and EU-MIDIS II (2016) indicates that hate-motivated harassment of respondents with North African background has increased.
Results differ significantly between countries. For example, the rates of hate-motivated harassment experienced by respondents with Sub-Saharan African background vary widely across EU Member States. This suggests that EU Member States can potentially learn from each other’s experiences in terms of measures to prevent hate-motivated harassment.
A larger share of second-generation respondents than first-generation respondents experienced hate-moti-vated harassment in the 12 months before the survey (32 % and 21 %, respectively). Second-generation respondents also more often experienced recurring incidents of hate-motivated harassment.
Most respondents (81 %) who experienced harass-ment of any kind felt that one or more of these inci-dents was motivated by their ethnic or immigrant background. Furthermore, 57 % of those who said they experienced a physical attack indicated that one or more incidents were motivated by their ethnic or immigrant background.
Most respondents experienced hate-motivated harassment in the form of offensive or threatening comments or gestures in person. Experiences with personal cyber-harassment motivated by hatred were less common. Young respondents experienced this more often than older ones. FRA’s EU-wide survey on violence against women also found young women to be at a higher risk of cyber-harassment than women from other age groups.
Most respondents describe perpetrators of hate-motivated harassment (71 %) and violence (64 %) as someone from the ‘majority population’. However, 23 % of victims of hate-motivated harassment say the perpetrator was from another ethnic minority, and 8 % say that the perpetrator had the same ethnic or immi-grant background as themselves. This proportion was much higher for respondents with Sub-Saharan African
17
Key findings and FRA opinions
background in France (35 %) and Sweden (44 %), Roma in Bulgaria (42 %) and Romania (40 %), as well as respondents with Asian background in Cyprus (45 %). Respondents often did not know the perpetrators of the hate-motivated harassment (72 %) and violence (49 %) they experienced. However, incidents of hate-motivated violence against women were more often perpetrated by someone known to the victim, such as a neighbour.
Results suggest that the overall prevalence of hate-motivated harassment and violence is similar for women and men. However, their experiences may differ in other ways, such as in terms of the perpe-trators. That incidents against women more often involve somebody they encounter in their everyday lives – such as an acquaintance or neighbour – may make some of the incidents particularly difficult for the victim to report. Indeed, women also indicated more often than men that they did not report an incident of hate-motivated violence because they were afraid of intimidation or retaliation from the perpetrators.
Among respondents who experienced hate-moti-vated harassment, 3 % said that the perpetrator was a police officer or a border guard, and 4 % indicated that the perpetrator was a public official. A very small proportion (1 %) of respondents said that they had been physically assaulted by a police officer – such as pushed, hit or kicked – because of their ethnic or immigrant background in the 12 months before the survey. Nevertheless, this can potentially affect eth-nic minority and immigrant communities’ trust in the police, who have a duty to protect human rights.
In light of these results, efforts to prevent hate-motivated harassment should consider that most of these incidents happen in the street and in other public places, rather than online. This being said, the higher prevalence of hate-motivated cyber-harass-ment among young respondents, as well as the need to find ways to make it easier for victims to report incidents, call for considering new ways of report-ing incidents – for example, via online tools. In this regard, the Commission’s #NoPlace4Hate initiative to implement a code of conduct on countering illegal hate speech online, agreed to by Facebook, Twitter, Microsoft and YouTube in May 2016, has had some notable results. On average, one year later, IT compa-nies have responded to notifications concerning illegal hate speech by removing the content in 59 % of cases. This is more than twice the level of 28 % recorded six months earlier. In addition, the amount of notifications reviewed within 24 hours improved from 40 % to 51 % in the same six-month period.15
15 European Commission (2017).
The overwhelming majority (90 %) of respondents who were asked about the most recently experienced incident of hate-motivated harassment indicated that they did not report the incident – either to the police or to another organisation or service. Of those who did report such incidents, 36 % reported them to the police, 53 % to another organisation/service, and 10 % to both the police and another organisation/service. Only 13 respondents reported hate-motivated har-assment to an equality body, human rights insti-tution or ombudsperson (out of 8,709 respondents who provided details of the most recent incident they had experienced).
Compared with incidents of hate-motivated harass-ment, respondents who experienced hate-motivated physical attacks were more likely to report these inci-dents. Overall, 28 % reported the most recent incident of hate-motivated violence in the five years before the survey to the police or to another organisation or ser-vice. Incidents of violence were most often reported to the police, while harassment was usually reported to some other organisation or service – for example, somebody at the place where it occurred.
When asked why they decided not to report an inci-dent of hate-motivated violence, respondents most often noted that they were not convinced that any-thing would happen or change as a result of report-ing (41 %). Other common reasons for not reporting included wanting to deal with the problem oneself or with the help of family and friends (21 %), and the per-ception that the incident was minor and therefore not worth reporting (16 %). Furthermore, 11 % mentioned not trusting the police or being afraid of the police.
These results point to a risk that measures such as individual victims’ needs assessments, as provided for in the Victims’ Rights Directive, would benefit only the very small minority of victims who do report incidents. While some victims of hate-motivated incidents may find other meaningful ways of coping with the experi-ence – such as relying on assistance from family and friends – some of them might have chosen to seek legal redress if they had access to more information about their rights and existing support mechanisms. When crime is not reported to authorities, this in most cases also means that the incidents are never officially investigated and perpetrators are not prosecuted. The finding that many victims do not report victimisation to the authorities is consistent with findings of other victimisation surveys and FRA’s research on specific groups, such as lesbian, gay, bisexual and transgender persons, Jewish people, and women.
Second European Union Minorities and Discrimination Survey – Main results
18
FRA opinion 5 EU Member States should encourage victims to report hate crime. This can include initiatives that make it easier for victims to report incidents, such as online reporting tools. In addition, Member States should consider allowing for third-party reporting to ensure that more incidents – reported to a range of civil society organisations – come to the attention of criminal justice. Such cooperation can foster confidence in law enforcement and the legal system among immigrant and ethnic minority groups.
EU Member States should ensure that any violent incident involving law enforcement is investigated independently through fully independent and functional complaints mechanisms.
FRA opinion 6EU Member States should ensure that hate crime incidents are recorded more effectively by law enforcement. This could be achieved by ensuring that law enforcement and criminal justice adopt the right tools and methods in consultation with civil society. This includes establishing a bias motivation using a range of indicators – including the victim’s perception, or that of any witness, that the offence was motivated by bias, prejudice or hostility. In this respect, Member States can usefully draw from work published by the OSCE/ODIHR in ‘Hate Crime Data Collection and Monitoring: A Practical Guide’. Frontline law enforcement and criminal justice officials should be systematically trained in identifying, recording, investigating and prosecuting hate-motivated crime effectively.
FRA opinion 7EU Member States should strengthen their victim support services in line with the Victims’ Rights Directive, which makes explicit reference to victims of hate crime.
1�3� Police stops and treatment by the police
Overall, 14 % of respondents were stopped by the police in the 12 months preceding the survey (including stops in a private vehicle, stop-and-search incidents on the street, or in public transport). Of those stopped, 40 % believe that the most recent stop was because of their immigrant or ethnic minority background. Respondents with Asian and South Asian, North Afri-can, and Sub-Saharan African backgrounds, as well as Roma, more frequently say that they were stopped by the police because of their immigrant or ethnic minor-ity background. None of the respondents from the Rus-sian minorities indicated that they were stopped by the police because of their minority background. On aver-age, the police stopped young persons with immigrant
backgrounds more often than older persons, and men more often than women, across most target groups surveyed. Among Roma, the rates of police stops are quite similar across different age groups. Roma women and men also believe to the same extent that the most recent police stop was because of their minority background. These results indicate that discriminatory police practices affect certain immigrant and ethnic minority groups more than others, confirming similar findings in EU-MIDIS I.
Law enforcement has a duty to treat everyone respectfully, addressing the needs and rights of all victims. Discriminatory ethnic profiling – police stops based solely or mainly on an individual’s personal characteristics rather than on their behaviour – is unlawful. It can also damage community relations and undermine trust in, and public cooperation with, law enforcement. EU-MIDIS II results show that most respondents generally trust the police and say that they were respectfully treated during the most recent police stop. However, the reporting rates for hate crime incidents remain very low, which could indicate a lack of confidence in law enforcement’s ability to tackle hate crime incidents effectively.
FRA opinion 8EU Member States should stop unlawful discrimi-natory ethnic profiling by law enforcement. They should raise awareness among police forces of the damaging effect of such practices on community relations and trust in law enforcement.
1�4� Living together in the EU: citizenship, participation, trust and tolerance
EU-MIDIS II findings highlight that experiences of dis-crimination and victimisation can affect social cohesion in European societies by showing that such experi-ences affect respondents’ levels of attachment to the countries they live in and, most importantly, their trust in public institutions. FRA’s EU-MIDIS II report on Selected Findings on Roma also extensively discusses the damaging consequences of discrimination and vic-timisation experiences on social inclusion.
The integration of third-country nationals falls under EU Member States’ competence. However, integra-tion matters are strongly interlinked with the prin-ciple of non-discrimination and so also with the EU’s equality framework. The EU supports national and local policies on integration through policy coordi-nation, knowledge exchanges and financial support, based on TFEU Article 79(4). Fostering the integration of migrants also promotes various principles of the
19
Key findings and FRA opinions
European Pillar of Social Rights, while efforts on the pillar also have the potential to enhance integration. The Common Basic Principles for immigrant integration policy (2004)16 were essential for Member States to agree on shared goals on integration and in assisting Member States in formulating integration policies. In 2016, the European Commission launched an Action Plan presenting a framework for action and concrete initiatives to support Member States in the integration of non-EU nationals residing legally in the EU.17 The action plan includes actions to support pre-departure and pre-arrival measures, education, employment and vocational training, access to basic services, active participation and social inclusion.
The Urban Agenda for the EU, an initiative aiming to create new forms of cooperation on urban issues, provides a framework for better coordination at EU, national and regional/local levels in improving the design and implementation of policies. The Partner-ship on Inclusion of Migrants and Refugees, created within the Urban Agenda framework in 2016, is a step towards this direction.
Compared to Roma or other ethnic groups without migration background, immigrants face additional challenges with respect to societal participation. These challenges are very often related to their resi-dence status, their citizenship, or are linked to the pro-cess of family reunification. EU-MIDIS II results show that, while the majority of immigrants hold a secure residence status and often hold national citizenship, a considerable share does not have long-term resi-dence rights or citizenship – even after many years of residing in a country. Meanwhile, almost all descend-ants of immigrants hold a secure residence status – though many do not obtain citizenship of the country they live in.
Having a secure residence status and particularly acquiring citizenship can promote integration and strengthen individuals’ sense of attachment to their country of residence. Having an insecure residence status can be particularly problematic for vulnerable groups, such as victims of violence. This is recognised, for example, in the Council of Europe Convention on preventing and combating violence against women and domestic violence, which addresses the need for victims of domestic violence – who are often depend-ent on their spouses – to obtain secure residence status (Article 59). Victims without a secure residence status may also be more reluctant to contact or report to the competent authorities in case of discrimination, harassment or hate-motivated violence.
16 Council of the European Union and Justice and Home Affairs (2004).
17 European Commission (2016).
FRA opinion 9EU Member States should take into account the potential positive impact of secure residence status on integration when reviewing their national immigration legislation. They should consider removing potentially unnecessary obstacles to accessing long-term residence status or national citizenship.
The results show that immigrants often indicate higher levels of trust in national and European institu-tions than the general population does. However, the results vary considerably across countries and target groups. Overall, trust in local authorities, the police and the legal system is particularly high, though lower among the second generation in most countries. As previously noted, the survey findings indicate that experiences of discrimination and hate crime victimi-sation have a strong negative impact on respondents’ levels of trust in public institutions and their feelings of belonging to, and identifying with, the country they live in. Both acquiring citizenship of the country of residence and being treated equally strengthen individuals’ identification with their country of resi-dence, and so ultimately bolster social cohesion.
FRA opinion 10EU Member States should place anti-discrimination measures at the core of their national integration policies, in line with the Common Basic Principles for Immigrant Integration Policy in the EU. Furthermore, EU Member States should strengthen measures that improve mutual understanding, participation and trust between immigrants and host societies.
EU-MIDIS II shows that respondents’ educational level and their proficiency in at least one of the national languages of the country they live in are positively related to their position in the labour market. This is especially true for the second generation. Access to education and employment are critical elements for successful integration. According to the Common Basic Principles for immigrant integration, basic knowledge of the host society’s language, history, and institutions is indispensable to integration, and efforts in education are critical to preparing immigrants – and particularly their descendants – to be more successful and more active participants in society. However, the data also suggest that a large proportion of respondents can-not read in the respective national language. In the EU-28, on average, 74 % of the population aged 15 to 64 had completed at least upper secondary education in 2016. Meanwhile, only 61 % of EU-MIDIS II respond-ents aged 16 to 64 reported having completed at least upper secondary education.
Second European Union Minorities and Discrimination Survey – Main results
20
FRA opinion 11EU Member States should take measures to increase the participation of ethnic minority children and children of immigrants in early childhood education and care, which is important for later educational achievement. This would help achieve the Europe 2020 target of ensuring that at least 95 % of all children aged between four and the starting age of compulsory education participate in early childhood education and care. In addition, EU Member States should adequately address early school leaving, the rate of which is markedly higher for Roma and children born outside the EU.
EU Member States should ensure that immigrants and members of ethnic minorities who are not native speakers have access to free-of-charge general and job-related language(s) training to improve their labour market participation and enhance their overall social integration. In addition, Member States should promote recruitment of teachers with ethnic minority or immigrant backgrounds and provide teachers with training on diversity and non-discrimination.
FRA opinion 12EU Member States should increase their efforts to provide children with ethnic minority or immigrant backgrounds with specific learning support at school so that they can complete at least upper secondary education to improve their chances in the employment field. This would contribute to the Europe 2020 strategy target to reduce the share of early leavers of education and training. Furthermore, it would contribute to reaching targets of UN Sustainable Development Goal 4, which aims to ensure inclusive and equitable quality education and promote lifelong learning opportunities for all.
21
2 What do the results show?
2�1� Discrimination and awareness of rights
KEY FINDINGS
n One in four (24 %) respondents felt discriminated against because of their ethnic or immigrant background in the 12 months preceding the survey, a small drop compared to EU-MIDIS I, where one in three respondents (30 %) stated that they felt discriminated against because of their ethnicity in the past 12 months.
n When looking at the past five years, four out of 10 respondents (38 %) felt discriminated against because of their ethnic or immigrant background, skin colour or religion in daily life.
n Similarly to the findings of EU-MIDIS I, ‘ethnic origin or immigrant background’ emerges as the most common ground of discrimination – experienced by every fourth respondent (25 %) in the five years preceding the sur-vey, followed by skin colour and religion (each 12 %).
n The second generation experiences higher levels of religious discrimination than the first generation (20 % versus 12 %).
n Among the groups surveyed, similarly to EU-MIDIS I, Roma respondents (26 %) and individuals from North Africa (31 %) and Sub-Saharan Africa (24 %) indicate the highest levels of discrimination based on ethnic or immigrant background in the 12 months before the survey.
n The highest rates of discrimination based on ethnic or immigrant background are observed in the area of employment and when accessing public and private services. Across all areas of life examined by the survey, almost a third (29 %) of respondents who looked for work in the five years preceding the survey felt discrimi-nated against, and one in 10 (12 %) experienced this in the year before the survey.
n Respondents cite their skin colour or physical appearance and their first or last name as the main reasons for experiencing discrimination in almost all areas of life.
n Only one out of eight respondents (12 %) reported or filed a complaint about the most recent incident of discrimination they experienced because of their ethnic or immigrant background. Women reported slightly more often than men (14 % and 11 %, respectively). In the 2008 EU-MIDIS I survey, the overall reporting rate was 18 %. This means that incidents of discrimination continue to remain largely unreported – and therefore invisible to institutions with a legal obligation to respond.
Second European Union Minorities and Discrimination Survey – Main results
22
Article 21 of the EU Charter of Fundamental Rights pro-hibits discrimination based on racial or ethnic origin. Since 2000, the EU has put in place robust legislation on the issue. This includes the Racial Equality Directive, which offers protection against discrimination based on racial or ethnic origin. In addition, the Employment Equality Directive prohibits discrimination based on reli-gion or belief in employment and occupation. However, 17 years after adoption of these directives, the results of this survey underscore that many people with immi-grant or ethnic minority backgrounds face discrimina-tion in a range of settings, and many still do not know their rights or where to turn to report such incidents.
The EU is committed to fulfilling the UN Sustainable Development Goals. Goal 10 aims “to reduce inequality within and among countries”, and Target 10.3 specifi-cally refers to ensuring “equal opportunity and reduce inequalities of outcome, including by eliminating dis-criminatory laws, policies and practices and promoting appropriate legislation, policies and action in this regard”. The data of this survey can be used to populate Indicator 10.3.1: “Proportion of the population reporting having personally felt discriminated against or harassed within the previous 12 months on the basis of a ground of dis-crimination prohibited under international human rights law”. This same indicator is also used to gauge progress on Target 16.b (“Promote and enforce non-discriminatory laws and policies for sustainable development”) of Goal 16, which aims to “promote peaceful and inclusive soci-eties and to provide access to justice for all and build effective, accountable institutions at all levels”.18
2�1�1� Discrimination experiences
Measuring discrimination in EU-MIDIS II
The survey asked respondents if they had felt dis-criminated against on one or more grounds – skin colour, ethnic origin or immigrant background, reli-gion or religious beliefs, sex, age, disability, sexual
18 See the UN’s webpage on the SDG Indicators.
orientation, and ‘other’ grounds – in different do-mains and activities:
• when looking for work,
• at work,
• in education or when in contact with their chil-dren’s school personnel,
• in access to healthcare,
• in connection with housing, and
• when using public or private services (such as public transport, administrative offices, when entering a night club, restaurant or a hotel, and when being in or entering a shop).
In this report, the discrimination rates specified in-dicate the percentage of respondents who felt dis-criminated against in at least one of the domains investigated. The rates are calculated for the 12- month and five-year periods preceding the survey. Determining the rates of discrimination based on the various individual grounds, which would make it possible to identify the most common ground of discrimination among the eight different grounds asked about in the survey, was only possible for four areas of life (when looking for work, at work, in access to housing, and when in contact with school authorities as a parent), and only with respect to the five years preceding the survey.
Respondents who indicated having experienced dis-crimination on at least one of three specific grounds – skin colour, ethnic origin or immigrant background, and religion or religious beliefs – were asked further details about the incident, applying the generic term ‘ethnic or immigrant background’. Results based on information covered in these subsequent questions, therefore, cannot be further disaggregated along the three individual grounds.
Discrimination on all grounds
As in EU-MIDIS I, this survey asked respondents about their experiences with discrimination based on a number of different grounds during the past five
n Reported incidents are mostly related to discrimination at the workplace (40 %) or in public services (22 %) and most complaints were made to an employer (36 %), trade unions and staff committees (13 %) or to the police when related to entering a night club or a bar (17 %). Only 4 % of all reports were made to an equality body.
n Satisfaction with the way complaints were handled is low overall, regardless of where they were filed. Re-spondents who did not report discrimination incidents to anyone said that they failed to do so because nothing would happen or change as a result of reporting.
n Most respondents (71 %) are not aware of any organisation that offers support or advice to discrimination victims and the majority (62 %) is not aware of any equality body. This could partly explain the low reporting rates.
n Regarding awareness of anti-discrimination legislation, a majority of respondents (67 %) know that discrimina-tion based on skin colour, ethnic origin or religion is unlawful in their country.
23
What do the results show?
years. Respondents could indicate up to eight differ-ent grounds of discrimination in each of four areas of life: when looking for work, at work, in access to housing and when in contact with school authorities (as a parent or guardian).19 Due to the survey ques-tionnaire’s overall length, different grounds of dis-crimination were only asked about for the five-year period preceding the survey. By contrast, respondents were also asked about the grounds ‘ethnic origin and immigrant background’, ‘skin colour’ and ‘religion or religious beliefs’ in connection with the 12-month period preceding the survey. However, these were combined under the generic term ‘ethnic or immigrant background’ (only respondents who felt discriminated against on at least one of these three grounds were asked follow-up questions). Therefore, discrimination rates based on different grounds can only be pre-sented for the five-year period.20
This section looks at the extent to which survey respondents felt discriminated against on the basis of different grounds in the five years preceding the survey. The different grounds of discrimination may individu-ally have prompted experiences of discrimination or
19 Different grounds of discrimination were also asked about in the area of health, but, due to a routing mistake, this domain cannot be considered for this analysis. Results for this domain are considered in the 12-month overall rate of discrimination based on ethnic or immigrant background. Multiple grounds were not asked about for the category ‘other public or private services’, which includes education, public transport, public administration, restaurant or bar, and shop.
20 Based on findings in EU-MIDIS I that ‘ethnic origin’, ‘skin colour’ and ‘religion or religious belief’ are the three most common grounds of discrimination, the survey only collected further, detailed information for these three grounds in combination.
may have combined with other grounds to aggravate individuals’ experiences of discrimination.
Figure 1 shows that, as in EU-MIDIS I, ‘ethnic origin or immigrant background’ is the most common ground for discrimination – experienced by every fourth sur-vey respondent (25 %) in the five years preceding the survey. 12 % of respondents felt discriminated against because of their skin colour and 12 % because of their religion or religious belief, followed by age (7 %), sex/gender (2 %), and disability (1 %).
The majority of respondents who felt discriminated against in one of the four areas mention only one ground for discrimination (20 %). However, every tenth respondent (11 %) indicates having felt discriminated against on two grounds in the previous five years, another 5 % say they suffered discrimination based on three different grounds, and a further 1 % of respond-ents indicate four or more grounds for discrimination in any of the four areas of life considered for the analysis.
The results also point to an intersection of religion and ethnic origin; 70 % of all respondents who indicate
Figure 1: Grounds of discrimination experienced in four areas of daily life in 5 years before the survey (%) a,b
0 402010 605030
50.212
71212
25
AgeGender
DisabilitySexual orientation
Ethnic origin
Religion
Other
Skin colour
Notes: a Out of all respondents at risk of discrimination on different grounds in at least one of four domains of daily life asked about in the survey (‘in 5 years before the survey’: n=21,541); weighted results.
b Domains of daily life considered for analysis: looking for work, at work, education (as parent or guardian) and housing.
Source: FRA, EU-MIDIS II 2016
Second European Union Minorities and Discrimination Survey – Main results
24
religion as a ground for discrimination also felt discrimi-nated against because of their ethnic origin or immi-grant background. By contrast, 36 % of all respondents who felt discriminated against because of their ethnic origin or immigrant background indicate that they also experienced discrimination based on religion – a find-ing that could indicate that many respondents perceive religion as a sub-dimension or a distinct element of their ethnic or immigrant background. Another reading of these results would note that there are more respond-ents who feel discriminated against because of their ethnic or immigrant background than respondents who feel discriminated against because of their religion. The results also show an interaction between the grounds ‘skin colour’ and ‘ethnic origin or immigrant background’ (63 % of respondents who mention skin colour also indicate ethnic origin or immigrant background). By contrast, the grounds ‘skin colour’ and ‘religion or reli-gious beliefs’ intersect to a lesser extent – 28 % of all respondents who mention skin colour as a ground for discrimination in the past five years also mention religion or religious beliefs.
With regard to the four areas of life, the results show that, on average, respondents feel most discrimi-nated against – and on more than one ground – when they look for work, and to a slightly lesser extent at their work place.
A look at gender differences reveals that more women than men indicate experiencing discrimination based on sex/gender (4 % and 1 %, respectively). With regard to other grounds, there are no substantial dif-ferences between men’s and women’s experiences – except that more men indicate encountering dis-crimination based on skin colour (14 % compared to 10 %). However, results with regard to gender differ substantially across target groups and countries, as shown further below.
There are some noteworthy differences between age groups with regard to grounds of discrimina-tion, as shown in Figure 2. For example, younger respondents experience religious discrimination to a higher extent than older respondents (16 % of 16–24 year-olds and 13 % of 25–44 year-olds, compared to 10 % of 45–59 year-olds and 7 % of 60+-year olds). Age discrimination is also mainly mentioned by the very youngest or very oldest age groups (14 % and 10 %, respectively). By contrast, the first and sec-ond most cited grounds – ‘ethnic origin or immigrant background’ and ‘skin colour’ – are relevant in all age groups, though to a different extent, as shown in Figure 2.
The differences between age groups are in a sense reflected in observed differences between first- and
Figure 2: Discrimination on different grounds in four areas of daily life in 5 years before the survey, by age group (%) a,b
0
40
20
60
10
50
30
60+45-5925-4416-24
Ethnic origin or immigrant background Religion or religious beliefsSkin colour Age
24
10
1614
27
14 13
5
23
12 106
13
7 7 10
Notes: a Out of all respondents at risk of discrimination on different grounds in at least one of four domains of daily life asked about in the survey (‘in 5 years before the survey’: n=21,541); weighted results.
b Domains of daily life considered for analysis: looking for work, at work, education (as parent or guardian) and housing.
Source: FRA, EU-MIDIS II 2016
25
What do the results show?
second-generation respondents21 in terms of feeling discriminated against on the grounds of ‘religion or reli-gious beliefs’, ‘ethnic origin or immigrant background’, and ‘age’. On average, one out of five second-gener-ation respondents (20 %) mention religious discrimi-nation, compared to one out of eight first-generation respondents (12 %). The difference between the second and first generation is slightly less pronounced regard-ing the grounds ethnic origin or immigrant background (29 % vs 24 %) and age (9 % vs 6 %).
Looking at different grounds of discrimination from the perspective of the aggregate target groups, which are based on respondents’ countries/regions of origin, the aggregated results in Figure 3 show that, for all target groups other than respondents from Sub-Saharan Africa and the Russian minority, ‘ethnic origin or immigrant
21 The differentiation between first- and second-generation respondents does not apply to ‘recent immigrants’, Roma, and Russian minority respondents. Second-generation respondents are on average younger (majority: 16 to 44 years old).
background’ is the main ground of discrimination. This ground is mentioned most often by respondents of North-African background (36 %) and Roma (27 %). By contrast, respondents from Sub-Saharan Africa indicate that they mostly experienced discrimination based on skin colour (27 %), which is also the second most relevant ground for discrimination experienced by Roma respondents (19 %). Religious discrimination is particularly pertinent for respondents from North Africa and Turkey (20 % and 16 %, respectively).22 Age dis-crimination mainly affects respondents of North-African (10 %), Sub-Saharan African (9 %) and Russian ethnic minority backgrounds (6 %). Figure 3 also shows that some target groups – for example, respondents from North Africa or Sub-Saharan Africa – on average experi-ence both higher levels of discrimination and greater variation with regard to grounds of discrimination.
22 As shown in the EU-MIDIS II report on Muslims – Selected findings, the overwhelming majority of EU-MIDIS II respondents with North African and Turkish backgrounds identify themselves as Muslims when asked about their religion. See FRA (2017b).
Figure 3: Grounds of discrimination experienced in four areas of daily life in 5 years before the survey, by target group (%) a,b,c,d
0
40
20
60
10
50
30
NOAFR ROMA TUR SSAFR RIMGR (S)ASIA RUSMIN
Ethnic origin or immigrant background Skin colourReligion or religious beliefs
GenderAge
36
20
9 10
4
27
1
19
41
1916
2 3 2
19
5
27
9
3
16
2 (1) (2) (1)
128 7
2 (1)5 6
(1)
Notes: a Out of all respondents at risk of discrimination on different grounds in at least one of four domains of daily life asked about in the survey (‘in 5 years before the survey’: n=21,541); weighted results, sorted by ‘ethnic origin or immigrant background’.
b Results based on a small number of responses are statistically less reliable. Thus, results based on 20 to 49 unweighted observations in a group total or based on cells with fewer than 20 unweighted observations are noted in parentheses. Results based on fewer than 20 unweighted observations in a group total are not published.
c Domains of daily life considered for analysis: looking for work, at work, education (as parent or guardian) and housing.
d Acronyms for target groups refer to immigrants from [country/region] and their descendants: TUR = Turkey, SSAFR = Sub-Saharan Africa, NOAFR = North Africa, (S)ASIA = South Asia and Asia, RIMGR = recent immigrants from non-EU countries, RUSMIN = Russian minority, ROMA = Roma minority.
Source: FRA, EU-MIDIS II 2016
Second European Union Minorities and Discrimination Survey – Main results
26
Different target groups offer diverse perspectives
This section takes a more detailed look at the results regarding different grounds of discrimination from the perspective of the various target groups across coun-tries in which they were surveyed (Figure 4). Refer-ences to country-specific effects (when more than one target group was surveyed in one country) or to dif-ferences between men and women are made where relevant and significant.
SSAFR Among immigrants from Sub-Saharan Africa and descendants of such immigrants (SSAFR), the 5-year dis-crimination rate based on skin colour varies, ranging from 53 % in Luxembourg to 14 % in the United Kingdom. Higher rates of discrimination based on skin colour are also indicated by this group in Austria (45 %), Germany and Italy (37 % each). By contrast, religious discrimina-tion is mainly relevant for this group in Denmark, where one in four (25 %) felt discriminated against on this basis in the five years before the survey, and in Sweden and Italy (16 % and 10 %, respectively). This specific result
Figure 4: Grounds of discrimination experienced in four areas of daily life in 5 years before the survey, by target group and Member State (%) a,b,c,d
0
40
20
80
60
10
50
30
100
90
70
AT DK FR IT MT PT SEDE FI IE LU UK Group average
45
(5)
30
(3)
37
(3)
25
6
27 2528
(5)
24
6
30
4
34
6
23
13
32
(2)
28
(1)
37
10
34
(5)
53
(1)
42
8
33
(3)
2023
7(1)
32
16
29
7
14
(3)8
5
27
5
19
9
SSAFRSkin colour Ethnic origin Religion Age
0
40
20
80
60
10
50
30
100
90
70
AT DE NL Group average
ES FR ITBE DK SE BE NL Group average
29
(1)
20
4
24
59
(2)
17
(2)
15
3
22
5
17
(4)
41
6
27
6
20
(5)7 8
19
2
16
3
32
9
23
7
28
9
(1)(0)
37
9
20
12
40
20
29
5
42
13
31
6
36
9
20
10
TUR NOAFR
Ethnic origin Skin colourReligion Age
27
What do the results show?
Figure 4 (continued)
0
40
20
80
60
10
50
30
100
90
70
CY IT Group average
SIEL UK PL Group average
20
4(3)
51
11
24
32
812
9 8612
87
17
(3)(2)
16
(1)
16
2(1)
S(ASIA) RIMGR
Ethnic origin Skin colour Religion
Ethnic origin Skin colour Age
0
40
20
80
60
10
50
30
100
90
70
BG CZ HR HU RO Group average
LVEL ES PT SK EE LT Group average
19
58
37
10
3944
(3)
19
35
(1)5
42
(1)
23 22
3
15
61
(2)(0)
23
(1)
13
24
6
39
27
4
19
7 8(2) (4) (5) 6 5 6
ROMA RUSMIN
Notes: a Out of all respondents at risk of discrimination on different grounds in at least one of four domains of daily life asked about in the survey (‘in 5 years before the survey’: n=21,541); weighted results, sorted by the most common ground for each target group.
b Results based on a small number of responses are statistically less reliable. Thus, results based on 20 to 49 unweighted observations in a group total or based on cells with fewer than 20 unweighted observations are noted in parentheses. Results based on fewer than 20 unweighted observations in a group total are not published.
c Domains of daily life considered for analysis: looking for work, at work, education (as parent or guardian) and housing.
d Acronyms for target groups refer to immigrants from [country/region] and their descendants: TUR = Turkey, SSAFR = Sub-Saharan Africa, NOAFR = North Africa, (S)ASIA = South Asia and Asia, RIMGR = recent immigrants from non-EU countries, RUSMIN = Russian minority, ROMA = Roma minority.
Source: FRA, EU-MIDIS II 2016
Second European Union Minorities and Discrimination Survey – Main results
28
is to an extent linked to respondents’ countries of ori-gin and their predominant religious denomination. For example, respondents from Sub-Saharan Africa in Den-mark are predominantly born in Somalia (see Table 8 in Annex II), where the major religion is Islam. In Austria, on the contrary, immigrants from Sub-Saharan Africa are mostly born in Nigeria, which is equally divided between Christianity and Islam).
Age discrimination mostly plays a role for immigrants from Sub-Saharan Africa and descendants of such immi-grants in France (13 %), although respondents with North-African background in France also indicate much higher rates of age discrimination (12 %) compared to other groups and countries. This particular finding indi-cates a country-specific effect with regard to age as a ground for discrimination. In addition, on average, 13 % of all respondents with Sub-Saharan African or North African backgrounds in France say they felt discrimi-nated against because of their age when looking for work in the five years before the survey. This shows that being considered too young or too old and having a Sub-Saharan African (or North-African) background may make a difference in the French labour market, particularly when looking for work.
Gender differences among immigrants from Sub-Saha-ran Africa and descendants of such immigrants can be observed in most of the Member States surveyed. The rates for discrimination based on skin colour and ethnic origin are higher for Sub-Saharan men than women. For example, this is the case in Austria (for skin colour, men: 50 %, women: 27 %; for ethnic origin, men: 34 %, women: 15 %). It is also the case in France (for skin col-our, men: 37 %, women: 31 %; for ethnic origin, men: 26 %, women: 21 %). Similarly, the difference also exists in Italy (for skin colour, men: 43 %, women: 26%; for ethnic origin, men: 37 %, women: 28 %).
TUR & NOAFR Immigrants and descendants of immi-grants from Turkey (TUR) mainly indicate two grounds of discrimination – ethnic origin or immigrant back-ground (ranging from 17 % in Germany to 41 % in the Netherlands) and religion or religious beliefs (ranging from 7 % in Sweden to 27 % in the Netherlands). For both of these grounds, the rates of discrimination indi-cated this group are the highest in the Netherlands. Similarly, the results for the second target group in the Netherlands – immigrants and descendants of immigrants from North Africa (NOAFR) – also point to a country-specific effect: the discrimination rates for both grounds are at the highest level compared to the other countries in which this group was surveyed. 42 % of respondents with North-African background in the Netherlands felt discriminated against because of their ethnic or immigrant background in the five years pre-ceding the survey, and 31 % because of their religion or religious beliefs.
Immigrants and descendants of immigrants from North Africa also feel discriminated against because of their skin colour. Respondents from this group claim the high-est discrimination rate for this specific ground in Italy, where every fifth (20 %) felt discriminated against based on skin colour in the five years preceding the survey.
Among immigrants and descendants of immigrants from Turkey, gender differences exist with regard to discrimination based on religion or religious belief in Austria, Belgium and Germany: female respondents indicate higher rates of religious discrimination than their male counterparts. By contrast, in the Netherlands, twice as many men than women from this group say they felt discriminated against because of their religion in the five years before the survey (36 % vs 16 %). This is not the case for respondents with North Afri-can background in the Netherlands; among this group, more women than men mention religious discrimina-tion (35 % vs 27 %). In the countries where both target groups (TUR and NOAFR) were surveyed, men indicate higher rates of discrimination based on ethnic origin or immigrant background than women – except in Denmark and Spain, where no gender differences are observed.
(S)ASIA & RIMGR ‘Ethnic origin or immigrant back-ground’ is the most common reason for discrimina-tion for respondents with Asian (ASIA) and South Asian (SASIA) background. Rates range from every second such respondent in Greece (51 %) to almost every tenth in the United Kingdom (9 %). Skin colour is the second most common basis for discrimination against respondents with South Asian background in Greece: every fourth (24 %) felt discriminated against on this basis during the five years before the survey. In Greece, every tenth respondent with South Asian background (11 %) experienced religious discrimina-tion, as well. In Italy, such respondents indicate high rates of discrimination based on ethnic origin or immi-grant background, with a third (32 %) pinpointing this as a ground of discrimination; 12 % also mention skin colour. By contrast, respondents with Asian back-ground in Cyprus primarily refer to ethnic origin or immigrant background, which every fifth respondent (20 %) there indicates as the most common ground of discrimination. This is also the most relevant ground for recent immigrants (RIMGR) in Poland and Slovenia (17 % and 16 %, respectively).
Gender differences can be observed for respondents with South Asian background in Italy, though only for the ground 'ethnic origin' more women than men mention this as a ground for encountering discrimina-tion in the five years preceding the survey (38 % and 30 %, respectively).
ROMA For Roma respondents, in the countries in which they were surveyed, discrimination rates based
29
What do the results show?
on ethnic origin are highest in Portugal (61 %), Greece (44 %) and Croatia (42 %), and lowest in Bulgaria (19 %). Roma respondents also feel discriminated against because of their skin colour. For example, in the Czech Republic and Slovakia, this is the most cited ground for encountering discrimination in the five years before the survey (39 % each). In Croatia, 23 % of Roma respond-ents felt discriminated against because of their skin colour during that timeframe.
In Bulgaria, the Czech Republic and Spain, no gender differences are observed. In other countries in which Roma were surveyed, more Roma men than women felt discriminated against based on ethnic origin or skin colour. However, in Portugal, more Roma women than men felt discriminated against because of their ethnic origin in the five years before the survey (63 % and 58 %, respectively).
RUSMIN Compared to all other groups, respondents from the Russian minority target group (RUSMIN) indicate the lowest discrimination rates and the lowest variation in such rates based on different grounds. In the three Bal-tic countries in which this group was surveyed, Russians in Estonia indicate the highest discrimination rates in the five years preceding the survey: 8 % mention age dis-crimination and 7 % felt discriminated against because of their ethnic origin. Age discrimination is also mentioned by 6 % of such respondents in Latvia. The lowest rates of discrimination are noted in Lithuania, where almost none of the respondents indicated having experiences with discrimination in the five years before the survey.
Discrimination based on ‘ethnic or immigrant background’
This section outlines further results relating to the most relevant ground for discrimination – ethnic or immigrant background. As previously noted, this incorporates the three grounds: ethnic origin or immigrant background, religion or religious beliefs, and skin colour. The survey collected more detailed information concerning such dis-crimination in up to 10 areas of life.23 The discrimination rates presented in this section relate to the 12 months preceding the survey. (By contrast, in the previous sec-tion, calculations were based on four areas of daily life and related to the five years preceding the survey).
On average, 38 % of all EU-MIDIS II respondents felt discriminated against because of their ethnic or immi-grant background in the five years before the survey; 24 % did so in the 12 months preceding the survey. In EU-MIDIS I, one in three respondents (30 %) stated that
23 Domains of daily life asked about in the survey: looking for work, at work, education (self or as parent), health, housing, and other public or private services including public administration, restaurant or bar, public transport, shop.
they felt discriminated against because of their ethnic-ity (with respect to nine areas of life) in the 12 months preceding that survey.
Discrimination based on ‘ethnic or immigrant background’ among target groups
Similarly to EU-MIDIS I, respondents with North Afri-can and Sub-Saharan African backgrounds and Roma respondents reported the highest levels of discrimi-nation based on ethnic or immigrant background (Figure 5). On average, every third respondent with North-African background (31 %), every fourth Roma respondent (26 %), and every fourth respondent with Sub-Sharan Africa background (24 %) felt discriminated against based on their ethnic or immigrant background in the 12 months preceding the survey. The 12-month discrimination rate for respondents with Turkish back-ground is slightly lower; however, one out of five respondents (20 %) still felt discriminated against due to their ethnic or immigrant background in one or more areas of daily life.
Comparing these results with those of EU-MIDIS I in 2008, the new data suggest a decrease in over-all discrimination experienced by persons of Sub-Saharan African background (2008: 41 %) and Roma respondents (2008 : 47 %) in the 12 months preced-ing the survey. Meanwhile, for other target groups interviewed in the two surveys – such as immigrants and descendants of immigrants from North Africa (2008: 36 %) and Turkey (2008: 23 %) – average lev-els of discrimination remained more or less the same. Although not always directly comparable, changes in the 12-month prevalence of perceived discrimination over time vary among the countries in which the same target groups were surveyed. The reasons underly-ing the variations need to be analysed in more depth – for example, by looking at how policies in place in Member States between the two surveys addressed discrimination against the groups regarding whom large differences are observed.
In EU-MIDIS II, there are some notable differences in the 12-month prevalence of discrimination towards different target groups and across Member States, ranging from 50 % to 4 % (Figure 5). The highest 12-month rates of discrimination based on ethnic or immigrant background are experienced by respondents with Sub-Saharan Afri-can backgrounds in Luxembourg and Finland (50 % and 45 %, respectively); with North African backgrounds in the Netherlands (49 %); and by Roma respondents in Greece and Portugal (48 % and 47 %, respectively). The lowest rates are experienced by respondents from the Russian minority in Estonia, Latvia and Lithuania (7 %, 6 %, and 4 %, respectively). A closer look at the preva-lence of discrimination for the same target groups in
Second European Union Minorities and Discrimination Survey – Main results
30
different Member States shows considerable variations across the Member States.
On average, there are no substantial differences between women’s and men’s experiences with discrimination based on ethnic or immigrant background during the 12 months preceding the survey (24 % and 23 %, respec-tively). However, there are some substantial gender differences within and across target groups – these are discussed in more detail below.
SSAFR High discrimination rates for respondents with Sub-Saharan African backgrounds are observed in Lux-embourg and Finland, where every second respondent from this target group felt discriminated against. They also experienced high discrimination in Austria (42 %) and Denmark (41 %). The lowest rates were noted in Portugal (17 %) and the United Kingdom (15 %).
In the 12 months preceding the survey, women of Sub-Saharan African descent experienced higher levels of discrimination than men of the same background in Den-mark (women: 45 %, men: 40 %), Finland (women: 48 %, men: 43 %), France (women: 31 %, men: 26 %), and Italy (women: 29 %, men: 19 %). By contrast, the 12-month dis-crimination rate for men of Sub-Saharan African descent is twice as high as that for women in Austria (men: 49 %, women: 21 %), and is also higher in Luxembourg (men: 54 %, women: 45 %), Portugal (men: 20 %, women: 14 %) and Sweden (men: 39 %, women: 36 %).
There are only a few notable differences in the experi-ences of first- and second-generation respondents from this target group, which may be because this group is on average younger and primarily composed of first-generation immigrants (see Table 7 in Annex II). For example, the 12-month discrimination rate for second-generation respondents with Sub-Saharan African back-ground in Portugal is more than twice as high as that for first-generation respondents (32 % vs 15 %). Second-generation respondents from this group also indicate higher discrimination rates in the United Kingdom (21 % vs 12 %), Luxembourg (54 % vs 48 %), and France (33 % vs 27 %). By contrast, in Finland, more first-generation than second-generation respondents from this group felt discriminated against because of their ethnic or immi-grant background (including skin colour and religion) in the 12 months before the survey (46 % vs 40 %).
TUR Immigrants and descendants of immigrants from Turkey feel most discriminated against in the Nether-lands (39 %) and least discriminated against in Germany (18 %). However, 28 % of the respondents from this tar-get group in Austria, and one out of four in Denmark (26 %), felt discriminated against, in the 12 months before the survey, because of their ethnic or immigrant back-ground (including religious discrimination) in at least one of the 10 areas of life covered.
The rate for women with Turkish background in Austria is more than twice as high as that for men (38 % vs 16 %). In Germany, women from this target group also experi-ence higher levels of discrimination than men (21 % vs 16 %). By contrast, in the Netherlands and Denmark, more men from this target group feel discriminated against than women do (48 % vs 29 % for the Nether-lands, and 29 % vs 23 % for Denmark).
There are some notable differences between first- and second-generation respondents in this target group. For example, the discrimination rate for second-genera-tion respondents with Turkish background in Belgium is more than twice as high as that for first-generation immigrants (28 % vs 13 %). A similar pattern can be observed in Denmark (35 % vs 21 %). By contrast, in Austria and the Netherlands, first-generation respond-ents on average felt more discriminated against in the 12 months before the survey than second-generation respondents (Austria: 30 % vs 23 %; the Netherlands: 43 % vs 32 %).
NOAFR Respondents of North-African descent perceive similar levels of discrimination in almost all countries in which they were surveyed. The highest rates were indi-cated in the Netherlands, where every second respond-ent felt discriminated against in the 12 months preceding the survey (49 %). The lowest rates were noted in Spain; however, even there, every fifth respondent indicated having experienced discrimination (21 %). In Italy, France and Belgium, every third respondent mentioned experi-encing discrimination during this timeframe (34 %, 31 % and 31 %, respectively).
In the 12 months preceding the survey, more men than women of North African descent experienced discrimi-nation based on ethnic or immigrant background in the Netherlands (men: 54 %, women: 44 %) and in Bel-gium (men: 33 %, women: 28 %). In Italy, more women said they felt discriminated against than men did (37 % vs 31 %).
A look at differences between first- and second-gen-eration respondents shows that, on average, second-generation respondents with North-African background are more likely to experience discrimination than first-generation respondents from this target group (42 % vs 25 %). For example, in Belgium, the 12-month rate for second-generation respondents is 39 %, while for the first generation it is 23 %. (These results are similar to those for individuals with Turkish background in the country). The same pattern is observed in France (43 % vs 23 %), and in the Netherlands (58 % vs 44 %).
(S)ASIA Respondents with South Asian background indicated feeling most discriminated against in Greece (37 %) and least discriminated against in the United Kingdom (8 %) during the same period.
31
What do the results show?
Figure 5: Overall discrimination based on ethnic or immigrant background in 12 months before the survey, by survey target group and country (%) a,b,c,d
4020 806010 5030 90700
EE
ES
SK
Group Average
NL
UK
FR
SE
LU
LV
Group Average
HU
EL
AT
Group Average
IT
MT
FI
LT
RO
NL
PT
DK
PT
DE
AT
Group Average
EL
BG
IT
HR
BE
PL
UK
IE
DK
Total
CY
BE
CZ
DE
Group Average
IT
Group Average
FR
ES
SE
SI
Group Average
TUR
NOAFR
(S)ASIA
RIMGR
ROMA
RUSMIN
EU-28
SSAFR
100
21
5045
4241
38
3330
2923
1715
24
392826
2019
1820
49
3721
157
4847
3735
3230
2121
1426
76
(4)6
24
10
16
108
343131
31
38
Notes: a Out of all respondents at risk of discrimination based on ethnic or immigrant background in at least one of the domains of daily life asked about in the survey (‘in 12 months before the survey’: n=25,403); weighted results, sorted by 12-month rate within each target group.
b Domains of daily life asked about in the survey: looking for work, at work, education (self or as parent), health, housing and other public or private services (public administration, restaurant or bar, public transport, shop).
c Results based on a small number of responses are statistically less reliable. Thus, results based on 20 to 49 unweighted observations in a group total or based on cells with fewer than 20 unweighted observations are noted in parentheses. Results based on fewer than 20 unweighted observations in a group total are not published.
d Acronyms for target groups refer to immigrants from [country/region] and their descendants: TUR = Turkey, SSAFR = Sub-Saharan Africa, NOAFR = North Africa, (S)ASIA = South Asia and Asia, RIMGR = recent immigrants from non-EU countries, RUSMIN = Russian minority, ROMA = Roma minority.
Source: FRA, EU-MIDIS II 2016
Second European Union Minorities and Discrimination Survey – Main results
32
Other than in Italy, where men of South Asian descent felt more discriminated against in the 12 months before the survey than women (23 % vs 17 %), there are no substantial differences between men and women in this target group. However, this finding might be influenced by the sample composition with regard to gender in some of the countries in which this group was surveyed (see Table 7 in Annex II).
RIMGR The discrimination rates indicated by recent immigrants in Poland are twice as high as those indi-cated in Slovenia (15 % and 7 %, respectively). However, it should be noted that recent immigrants in Poland and Slovenia constitute a highly heterogeneous group in terms of their countries of origin (see Table 8 in Annex II). There are no substantial differences between the discrimination rates noted by women and men in this target group.
ROMA Roma respondents indicated feeling most dis-criminated against in Greece and Portugal, where every second felt discriminated against in the year preced-ing the survey (48 % and 47 %, respectively). They indicated feeling least discriminated against in Bulgaria (14 %).24 In only three of the nine countries in which Roma were surveyed were the indicated discrimination rates below the average of 26 % for this target group.
Substantial differences with regard to gender can be observed in Croatia, where more Roma men than women said they felt discriminated against in the 12 months before the survey (men: 44 %, women: 31 %). There are also substantial gender differences in Portu-gal, where more Roma women than men felt discrimi-nated against (women: 52 %, men: 42 %).
There are no substantial differences between the dis-crimination rates perceived by the various age groups in this target group.
RUSMIN As previously noted, respondents from the Russian minority on average feel the least discriminated against based on ethnic origin. On average, slightly more women than men in this target group feel dis-criminated against (women: 8 %, men: 4 %).
Country-specific effects
As Figure 5 shows, in some of the Member States in which more than one target group were surveyed, the level of discrimination varies substantially between dif-ferent target groups in the same country. For example, in Austria, the prevalence of discrimination perceived by respondents with Sub-Saharan African background is 42 %, while for respondents with Turkish background it
24 For further results on Roma based on the EU-MIDIS II survey, see FRA (2016).
is 28 %. A similar pattern can be observed for respond-ents from these two groups in Denmark (41 % vs 26 %), Sweden (38 % vs 19 %) and Germany (33 % vs 18 %). The biggest difference between groups can be observed in Portugal, where the discrimination rates for Roma respondents and respondents with Sub-Saharan African background differ by 30 percent points (47 % vs 17 %). This finding indicates that the same context (country) can have different impacts on different eth-nic or immigrant groups. Immigrants and descendants of immigrants from Sub-Saharan Africa and Roma especially experience higher levels of discrimination compared to other groups surveyed in the same coun-try. Given that skin colour is one of the most, or the most, relevant ground for discrimination for these two target groups, this finding shows that visible signs of difference such as skin colour – or, as shown in the EU-MIDIS II report on Muslims,25 visible signs of religious denomination such as headscarves or traditional cloth-ing – can trigger particularly high levels of discrimination in some Member States.
Another indication of a country-specific effect can be seen in the Netherlands, where both target groups surveyed (respondents with North African and Turkish backgrounds) experienced high levels of discrimination in the 12 months preceding the survey (49 % and 39 %, respectively). As previously noted, apart from ethnic origin, in the Netherlands both of these target groups indicate religion or religious belief as the second most important reason for having felt discriminated against. Compared to all other Member States surveyed, the level of perceived religious discrimination is highest in the Netherlands (as shown in Figure 4).
In some Member States where respondents with North African and Sub-Saharan African backgrounds were surveyed, the levels of discrimination are quite similar for both groups. This points to their particular vulner-ability to discrimination – although, as shown later in this report, the most important reason for feeling dis-criminated against (e.g. skin colour, ethnic origin) may differ between the groups. In France, the discrimina-tion rates for both target groups are around 30 % (of North African background: 31 %, of Sub-Saharan African background: 29 %). In Italy, respondents from all three target groups surveyed indicated relatively high rates of discrimination, though respondents with North African background mentioned experiencing discrimination in the 12 months before the survey more often (34 %) than respondents with Sub-Saharan African (23 %) or with South Asian backgrounds (21 %).
25 FRA (2017b).
33
What do the results show?
Frequency of discrimination experiencesThe survey asked respondents how often, in the 12 months before the survey, they felt discriminated against because of their ethnic or immigrant back-ground in five domains of life: when looking for work, at work, when accessing health services, in connec-tion with housing, and when in contact with school authorities. On average, respondents who indicated having felt discriminated against noted a minimum of 4.6 incidents a year, which shows that discrimination is a recurring experience for many victims. (Mean values vary between 1.9 incidents for immigrants with South Asian background in Italy to 6.5 incidents for immigrants with Turkish background in Sweden).
The frequency of discriminatory incidents per year varies across the five areas of life. However, respond-ents regularly feel discriminated against at work and when looking for work (Figure 6 and Figure 7). Of the respondents who indicated having felt discriminated against because of their ethnic or immigrant back-ground at work, 9 % said they experienced it on a daily basis. Meanwhile, 13 % said they felt discrimi-nated against more than 10 times in the 12 months preceding the survey (Figure 6). Determining what prompts a respondent to perceive discrimination as a daily occurrence would require further research – this may result either from one incident that affects the person every day or from a number of incidents
Figure 6: Number of discrimination experiences based on ethnic or immigrant background at work in 12 months before the survey (%) a,b
23
42
9
13
9 5Once2 to 5 times6 to 10 timesMore than 10 timesAll the time (daily)No answer
Notes: a Out of all respondents who felt discriminated against at work based on ethnic or immigrant background in the 12 months before the survey (n=1,427); weighted results.
b Question: “How many times has this happened to you in the past 12 months when at work?”Source: FRA, EU-MIDIS II 2016
Figure 7: Number of discrimination experiences based on ethnic or immigrant background when looking for work in 12 months before the survey (%) a,b
29
46
8
105 2
Once2 to 5 times6 to 10 timesMore than 10 timesAll the time (daily)No answer
Notes: a Out of all respondents who felt discriminated against when looking for work based on ethnic or immigrant background in the 12 months before the survey (n=1,919); weighted results.
b Question: “How many times has this happened to you in the past 12 months when looking for work?”Source: FRA, EU-MIDIS II 2016
Second European Union Minorities and Discrimination Survey – Main results
34
that lead to a constant feeling of being discriminated against on a daily basis.
Discrimination in different areas of everyday life
This section looks at discrimination rates in the differ-ent domains of everyday life covered by the survey. As previously noted, EU-MIDIS II measures discrimination experiences based on ethnic or immigrant background – including skin colour, ethnic origin or immigrant background, and religion or religious beliefs – in up to 10 areas of daily life during two time periods (five years and 12 months before the survey).
As Figure 8 shows, respondents indicated that they encountered the highest rate of discrimination based on ethnic or immigrant background during the five years before the survey when they were looking for work and when accessing public and private services. Some 29 % of all respondents who looked for a job in the five years before the survey felt discriminated against on this basis; 12 % experienced this in the year before the survey. When in contact with public or private services – such as administrative offices, pub-lic transport or when accessing a shop, restaurant or
bar – 22 % and 16 % of respondents felt discriminated against during the five years and 12 months before the survey, respectively.
Figure 9 shows the 12-month discrimination rate based on ethnic or immigrant background for each area of daily life and for each aggregated target group. For all groups, the highest level of discrimination is seen in the area of public or private services – ranging from 21 % for respondents with North African background to 5 % for respondents from the Russian minority. As previously noted, however, the calculation of the discrimination rate in this specific area includes the discrimination rates from four different domains – when being in contact with administrative offices, when using public transport, when accessing a shop, and when accessing a restaurant or a bar – and this partly explains its relatively high numbers when compared to the rates in other areas of daily life.
Similarly to EU-MIDIS I, respondents from all target groups indicated experiencing high levels of discrimination when looking for work; Roma respondents and respondents with North African background seem particularly affected (in the 12 months preceding the survey: 16 % and 15 %, respectively). Discrimination in access to housing also
Figure 8: Discrimination based on ethnic or immigrant background in different areas of life in 12 months and 5 years before the survey (%) a,b,c
0 402010 605030
3824
Total
3Health
126
Education
237
Housing
229
At work
2912Looking
for work
2216Other public/
private services
Past 5 yearsPast 12 months
Notes: a Out of all respondents at risk of discrimination based on ethnic or immigrant background in the particular domain (total n: ‘in 5 years before the survey’: n=25,228; ‘in 12 months before the survey’: n=25,403); weighted results, sorted by 12-month rate.
b Domains of daily life summarised under ‘other public or private services’: public administration, restaurant or bar, public transport, shop.
c Discrimination experiences in ‘access to health care’ were asked about only for the past 12 months preceding the survey due to a routing mistake in the questionnaire.
Source: FRA, EU-MIDIS II 2016
35
What do the results show?
appears to mostly affect these two target groups: around every tenth Roma respondent and respondent of North African background (12 % and 9 %, respectively) felt dis-criminated against because of their ethnic or immigrant background when trying to rent or buy an apartment or a house (see also the EU-MIDIS II report on Roma26). Roma respondents in the Czech Republic indicate the highest level of discrimination based on ethnic origin in access to housing, with one out of four (25 %) noting such an experi-ence in the 12 months before the survey. Discrimination in access to housing was also relevant for 14 % of Roma who looked for housing in Spain in that period, for 13 % of respondents with North African background in Italy, and for respondents with Sub-Saharan African background in Malta (12 %), Luxembourg (11 %) and France (8 %). Simi-larly, 7 % of recent immigrant respondents in Poland said they felt discriminated against when trying to rent or buy an apartment or a house in the year before the survey.
26 FRA (2016).
Discrimination at the work place in the 12 months pre-ceding the survey was mostly mentioned by respond-ents with North African and Sub-Saharan African backgrounds (14 % and 9 %, respectively).
The results on discrimination in access to health in the 12 months preceding the survey show that, on average, this is mostly experienced by Roma respondents (8 %); the highest rates in this area are observed for Roma in Greece (20 %), Romania (12 %), Slovakia (11 %) and Croatia (10 %). Apart from 10 % of the respondents with Turkish background in the Netherlands and 9 % of the respondents with South Asian background in Greece, no other groups indicated having experiences with dis-crimination when accessing healthcare services in the 12 months before the survey.
A discussion of more detailed findings for the areas of life in which, on average, the rate of discrimination based on ethnic or immigrant background in the 12 months preced-ing the survey is particularly high – such as in employ-ment or when using public or private services – follows.
Figure 9: Discrimination based on ethnic or immigrant background in different areas of life in 12 months before the survey, by survey target group (%) a,b,c,d,e
0
40
20
60
10
50
30
RUSMINRIMGR(S)ASIATURSSAFRROMANOAFR
At work
Health
Other public/private services
Housing
Looking for work
Education
5(2) (1)(2)(0)
6 7 52 1 1
14
7 73
6 6
15
10 96 4 3
1916
5 7 812
21
15 149
73 5 3 4 (3)(4)
(1)
Notes: a Out of all respondents at risk of discrimination based on ethnic or immigrant background in the particular domain (‘in 12 months before the survey’: n=25,403); weighted results.
b Domains of daily life summarised under ‘other public or private services’: public administration, restaurant or bar, public transport, shop.
c Discrimination experiences in ‘access to health care’ were asked about only for the past 12 months preceding the survey.
d Results based on a small number of responses are statistically less reliable. Thus, results based on 20 to 49 unweighted observations in a group total or based on cells with fewer than 20 unweighted observations are noted in parentheses. Results based on fewer than 20 unweighted observations in a group total are not published.
e Acronyms for target groups refer to immigrants from [country/region] and their descendants: TUR = Turkey, SSAFR = Sub-Saharan Africa, NOAFR = North Africa, (S)ASIA = South Asia and Asia, RIMGR = recent immigrants from non-EU countries, RUSMIN = Russian minority, ROMA = Roma minority.
Source: FRA, EU-MIDIS II 2016
Second European Union Minorities and Discrimination Survey – Main results
36
As the first graph in Figure 10 shows, there are substan-tial differences in the 12-month prevalence of discrimi-nation based on ethnic or immigrant background when accessing or using public or private services across the various target groups and across the Member States in which the specific target groups were surveyed, when compared with the average level of discrimination for the group as a whole or with the total prevalence of discrimination in this specific area of life. For exam-ple, the highest levels of discrimination in this area are observed for Roma in Greece and Portugal, where 43 % and 38 % of respondents, respectively, indicate such experiences; the group average is 19 %. Among respondents with North African background, the highest rate is in the Netherlands, where 39 % of respondents indicate having felt discriminated against when using public or private services; meanwhile, the group aver-age is 21 %. Respondents with Sub-Saharan African background in Austria (36 %), Luxembourg (33 %), and Finland (32 %) also note higher rates in this area than the average rate for this target group (15 %) or the total average rate (16 %).
The second graph in Figure 10 shows the 12-month dis-crimination rates in the area of education, experienced either by the respondents themselves or as parents or guardians when in contact with school authorities. The findings show that the level of discrimination in this domain is high for Roma respondents in Croatia (15 %) and for both respondents with Turkish and North African backgrounds in the Netherlands (12 % and 11 %, respec-tively). The findings for the Netherlands indicate that both immigrants and descendants of immigrants from North Africa and Turkey face similar extents of discrimi-nation in this life domain, even though the Netherlands is known to be one of the countries with favourable equality laws and strong enforcement mechanisms.
Such an indication for a country-specific effect can be seen not only in the Netherlands but also in Greece and in Austria. In these countries, the respec-tive target groups surveyed – Roma respondents and respondents with South Asian background in Greece, and respondents with Sub-Saharan African and Turk-ish backgrounds in Austria – indicate higher 12-month discrimination rates in the domain ‘other public and private services’ than respondents from other target groups and countries. Meanwhile, all respondents in the United Kingdom, independent of their backgrounds – South Asian or Sub-Saharan African – indicate sub-stantially lower levels of discrimination in this particu-lar area when compared to their groups’ averages or to the total average of discrimination in the 12 months before the survey.
Apart from the domain ‘other public or private services’, EU-MIDIS II respondents indicate experiencing the most discrimination when looking for work, followed
by incidents experienced at the workplace. Figure 11 shows the 12-month discrimination rates in both areas of employment: looking for work and at work. Roma respondents in Portugal indicate the highest discrimi-nation rates, with almost every second (47 %) who looked for work in the 12 months preceding the survey having felt discriminated against because of their eth-nic origin. Roma respondents in Croatia and the Czech Republic also experience high levels of discrimination when looking for a job (29 % and 28 %, respectively). A comparison of the discrimination rates Roma expe-rience when looking for work and at the work place reveals that Roma face high levels of discrimination when trying to enter the labour market. However, once they have a job, their experiences with discrimination do not substantially differ from those of respondents from other target groups, and do not substantially dif-fer across the countries in which they were surveyed.
This cannot be said for respondents of North African or Sub-Saharan African backgrounds. These two target groups face comparatively high levels of discrimination when seeking work. Respondents with North African background in Italy (23 %), Belgium (21 %) and the Netherlands (20 %) and respondents with Sub-Saharan African background in Austria (26 %) and Malta (20 %) are most affected. However, they are also more likely to experience discrimination on the job, with respond-ents of Sub-Saharan African background in Luxembourg, Sweden, and Ireland indicating the highest discrimina-tion rates in the 12 months before the survey (21 %, 17 % and 17 %, respectively).
Main reasons for discrimination in different domains
When asked about the main reason underlying the most recently encountered incident of discrimination based on ethnic or immigrant background in different domains of everyday life, respondents could choose among eight different reasons, and were allowed to select several reasons (multiple response): respondent’s skin colour/physical appearance, their first or last name, the accent/the way they speak the survey country language, their way of dressing (such as wearing a headscarf or tur-ban), the reputation of the neighbourhood in which the respondent lives, the respondent’s citizenship, and his or her country of birth. The results in Figure 12 show that two specific reasons are particularly relevant in all domains – respondents’ skin colour or physical appearance and their first or last names. While Roma respondents and respondents with Sub-Saharan African background mostly noted encountering discrimination based on their physical appearance, immigrants and descendants of immigrants from North Africa and Tur-key mainly indicate experiencing discrimination based on their names.
37
What do the results show?
Figure 10: Discrimination based on ethnic or immigrant background in ‘other public/private services’ and in ‘education’ (self or as a parent or guardian) in 12 months before the survey, by survey target group and country (%) a,b,c,d
Other public/private services Education
0 040 4020 2060 6010 1050 5030 30
16 6
(1)(0)(0)
(3)
7(2)4(4)
7799
(4)
1
7
777
11
12
7(4)6677
4(2)(2)(4)(4)5(6)77(8)101111
(4)
0(1)(3)(4)
(1)(5)
(10)15
5
1910
151717
2327
3038
43
5
7
58
1429
2110
182021
39
14
14
1111
1921
29
159
14171818
222528
303233
36
6
(3)
55
(3)
EE LV
ES ES
CZ ES
Group Average Group Average
NL NL
UK CY
IE AT
DE SE
AT DK
LV EE
Group Average Group Average
RO BG
EL HR
AT BE
Group Average Group Average
FR IE
SE FI
LU LU
LT LT
HU RO
NL NL
PT EL
DK DK
PT PT
MT IT
FI DE
Group Average Group Average
EL EL
BG PT
FR IT
ES CZ
DE DE
PL SI
UK UK
IT FR
DK MT
Total Total
CY UK
BE FR
SK SK
SE SE
Group Average Group Average
IT IT
Group Average Group Average
IT BE
HR HU
BE AT
SI PL
Group Average Group Average
TUR TUR
NOAFR NOAFR
(S)ASIA (S)ASIA
RIMGR RIMGR
ROMA ROMA
RUSMIN RUSMIN
EU-28 EU-28
SSAFR SSAFR
Notes: a Out of all respondents at risk of discrimination based on ethnic or immigrant background in the particular domain (‘in 12 months before the survey’, other public/private services: n=24,107; education: n=12,947); weighted results, sorted by 12-month rate within each target group.
b Domains of daily life summarised under ‘other public or private services’: public administration, restaurant or bar, public transport, shop.
c Results based on a small number of responses are statistically less reliable. Thus, results based on 20 to 49 unweighted observations in a group total or based on cells with fewer than 20 unweighted observations are noted in parentheses. Results based on fewer than 20 unweighted observations in a group total are not published.
d Acronyms for target groups refer to immigrants from [country/region] and their descendants: TUR = Turkey, SSAFR = Sub-Saharan Africa, NOAFR = North Africa, (S)ASIA = South Asia and Asia, RIMGR = recent immigrants from non-EU countries, RUSMIN = Russian minority, ROMA = Roma minority.
Source: FRA, EU-MIDIS II 2016
Second European Union Minorities and Discrimination Survey – Main results
38
Figure 11: Discrimination based on ethnic or immigrant background when ‘looking for work’ and ‘at work’ in 12 months before the survey, by survey target group and country (%) a,b,c
Looking for work At work
0 040 4020 2060 6010 1050 5030 30
12 9
5345566
(9)
4
5
14
101213
15
14
6677910
9(4)57
1010121415151717
21
(5)
57910
(3)6
1011
(2)
167
101313
2224
2829
47
(3)
(6)
(6)(8)
1114
1510
15202123
7
9
6(7)
131719
106(8)91011
151616181920
26
7
(1)
(2)(4)
(0)
EE LT
ES ES
HU ES
Group Average Group Average
NL NL
UK UK
DE AT
IE DK
AT LU
LV EE
Group Average Group Average
ES SK
PT PT
BE DK
Group Average Group Average
IT UK
FI DE
MT SE
LT LV
RO HU
IT FR
HR EL
AT SE
PT IT
DK FI
SE IE
Group Average Group Average
EL EL
BG BG
BE NL
CZ HR
SE AT
PL PL
UK PT
FR FR
LU MT
Total Total
CY IT
FR BE
SK CZ
DE DE
Group Average Group Average
IT CY
Group Average Group Average
NL IT
EL RO
DK BE
SI SI
Group Average Group Average
TUR TUR
NOAFR NOAFR
(S)ASIA (S)ASIA
RIMGR RIMGR
ROMA ROMA
RUSMIN RUSMIN
EU-28 EU-28
SSAFR SSAFR
Notes: a Out of all respondents at risk of discrimination based on ethnic or immigrant background in the particular domain (‘in 12 months before the survey’, looking for work: n=13,406; at work: n=16,836); weighted results, sorted by 12-month rate within each target group.
b Results based on a small number of responses are statistically less reliable. Thus, results based on 20 to 49 unweighted observations in a group total or based on cells with fewer than 20 unweighted observations are noted in parentheses. Results based on fewer than 20 unweighted observations in a group total are not published.
c Acronyms for target groups refer to immigrants from [country/region] and their descendants: TUR = Turkey, SSAFR = Sub-Saharan Africa, NOAFR = North Africa, (S)ASIA = South Asia and Asia, RIMGR = recent immigrants from non-EU countries, RUSMIN = Russian minority, ROMA = Roma minority.
Source: FRA, EU-MIDIS II 2016
39
What do the results show?
A summary of the findings regarding the differ-ent reasons for discrimination, including their rele-vance in different life domains and for the different target groups, follows.
Skin colour or physical appearance and respondents’ first or last names Skin colour or physical appearance is most frequently mentioned by respondents as a rea-son for discrimination regarding the use of healthcare services (59 %), at the work place (53 %), and when looking for work (50 %). First or last names are most often perceived as a reason for discrimination when respondents look for housing (44 %) or for work (36 %).
The findings at the level of the aggregate target groups show a clear pattern that is valid for four out of five areas of life: Roma respondents and immigrants and descendants of immigrants from Sub-Saharan Africa most often mention skin colour or physical appear-ance as the main reason for the latest incident of dis-crimination. For example, 81 % of Roma respondents who looked for work, 76 % of Roma respondents who looked for housing, and 72 % of Roma respondents who were in contact with their children’s schools indi-cated this reason. Similarly, among all respondents with Sub-Saharan African background, 82 % of those who looked for work, 84 % of those who looked for hous-ing, and 77 % of those in contact with their children’s schools noted skin colour as the main reason for the most recently encountered incident of discrimination in these contexts. Immigrants and descendants of immi-grants from South Asia and Asia also perceived skin col-our or physical appearance as a reason for encountering discrimination : 64 % indicated it as the main reason for encountering discrimination when looking for work, and 47 % for when looking for housing.
The other distinct cluster consists of immigrants and descendants of immigrants from North Africa and of immigrants and descendants of immigrants from Tur-key. For these two target groups, a person’s first or last name is more relevant than their skin colour or physical appearance. However, when it comes to the work environment, immigrants and descendants of immigrants from North Africa mention skin colour or physical appearance (45 %) in higher proportions than their first and last name (37 %). The same tendency can be observed in the area of healthcare: 54 % of this target group mention skin colour, while 34 % mention the first or last name.
When it comes to differences between generations, first or last names are perceived as more relevant by second-generation respondents (51 %) than by first-generation respondents (34%) in the context of looking for work. A similar pattern can be seen in the context of looking for housing: while 64 % of second-generation respondents who experienced discrimination in this area mentioned
their first or last names as the main reason, 38 % of first-generation respondents did so. The difference between generations is most significant for immigrants and descendants of immigrants from North Africa: in looking for work, 40 % of first-generation respondents vs 61 % of second-generation respondents; in looking for housing, 45 % of first-generation respondents vs 70 % of second-generation respondents.
Respondents’ addresses or the reputation of the neigh-bourhoods in which they live Notably, all eight rea-sons were mentioned by EU-MIDIS II respondents who looked for work (Figure 12). However, every ninth (12 %) respondent who felt discriminated against because of their ethnic or immigrant background when looking for work mentioned the reputation of their neighbourhood or residential address as the main reason for the most recent incident of discrimination – a factor pointed to this often only in the context of this particular domain. Furthermore, this specific reason was primarily noted by immigrants and descendants of immigrants from North Africa and Roma respondents. Among respondents with North African background, those living in France noted this reason at a higher percentage than respondents from the same target group in other Member States: 94 % of all respondents with North-African background who believe the address at which they live was the main reason for encountering discrimination when look-ing for work live in France. Meanwhile, every fifth Roma respondent in Greece (20 %), Slovakia (21 %) and Spain (18 %) identified the reputation of the neighbourhood or the address at which they live as the main reason for the latest incident of discrimination when looking for work.
The same is true for Roma respondents at the work-place and in healthcare. In these two domains, 16 % and 19 % of Roma respondents, respectively, mentioned this reason – while almost none of the respondents from the other target groups did so. Roma respondents in Spain (29 %), Croatia (27 %) and Slovakia (28 %) especially said, in much higher proportions than Roma in other countries, that the address at which they live triggered discrimination against them when accessing healthcare services.
Respondents’ citizenship and country of birth Respond-ents’ citizenship is ranked as the third reason underly-ing discrimination against respondents when accessing housing and when in contact with school authorities as parents or guardians. In these two contexts, it is cited as a reason for discrimination by every fifth respondent (22 % and 19 %, respectively). However, if respondents’ country of birth is considered in addition to their citi-zenship, the number of respondents who feel discrimi-nated against because of their ‘nationality’ or ‘origin’ substantially increases, particularly in the context of looking for housing or work or when in contact with school authorities as parents or guardians.
Second European Union Minorities and Discrimination Survey – Main results
40
Figure 12: Main reason for last incident of discrimination because of skin colour, ethnic origin or religion in 5 years before the survey, in five areas (multiple response) (%) a,b,c
Looking for workMy skin colour/my physical appearance
My first or last name
My accent/the way I speak (the language of the survey country)
The way I am dressed (such as wearing a headscarf/turban)
The reputation of the neighbourhood where I live (my address)
My citizenship
My country of birth
Other reason
50
36
18
14
12
13
7
4020 806010 5030 700
17
My accent/the way I speak (the language of the survey country)
The way I am dressed (such as wearing a headscarf/turban)
The reputation of the neighbourhood where I live (my address)
At workMy skin colour/my physical appearance
My first or last name
My citizenship
My country of birth
Other reason
53
25
15
10
3
13
13
4020 806010 5030 700
14
My accent/the way I speak (the language of the survey country)
The way I am dressed (such as wearing a headscarf/turban)
The reputation of the neighbourhood where I live (my address)
When using healthcare servicesMy skin colour/my physical appearance
My first or last name
My citizenship
My country of birth
Other reason
59
27
20
13
6
10
8
4020 806010 5030 700
12
41
What do the results show?
Figure 12 (continued)
My accent/the way I speak (the language of the survey country)
The way I am dressed (such as wearing a headscarf/turban)
The reputation of the neighbourhood where I live (my address)
When trying to rent or buy an apartment or houseMy skin colour/my physical appearance
My first or last name
My citizenship
My country of birth
Other reason
40
44
13
8
4
20
7
4020 806010 5030 700
22
My accent/the way I speak (the language of the survey country)
The way I am dressed (such as wearing a headscarf/turban)
The reputation of the neighbourhood where I live (my address)
When in contact with school authorities as a parent or a guardianMy skin colour/my physical appearance
My first or last name
My citizenship
My country of birth
39
28
19
9
5
11
4020 806010 5030 700
Other reason (14)
19
Notes: a Out of all respondents, who felt discriminated against because of their ethnic or immigrant background, in the different domains of daily life, (‘looking for work’: n=3,997; ‘at work’: n=2,993; ‘healthcare’: n=1,003; ‘housing’: n=1,507; ‘school authorities’: n=640); weighted results.
b Results based on a small number of responses are statistically less reliable. Thus, results based on 20 to 49 unweighted observations in a group total or based on cells with fewer than 20 unweighted observations are noted in parentheses. Results based on fewer than 20 unweighted observations in a group total are not published.
C Question: “Last time you felt discriminated against because of your ethnic or immigrant background when [domain], in your opinion, what were the main reasons for this?”
Source: FRA, EU-MIDIS II 2016
Second European Union Minorities and Discrimination Survey – Main results
42
Looking at results at the level of the aggregate target groups in the domain of housing shows that country of birth and citizenship are mostly identified as reasons for discrimination by respondents with South Asian or Asian backgrounds (citizenship: 37 %, country of birth: 35 %); by respondents with Turkish background (citizenship: 23 %, country of birth: 30 %); and by respondents with North African background (citizenship: 27 %, country of birth: 24 %). Persons’ nationality – their citizenship or country of birth – forms a barrier to entering the labour market particularly for respondents with Turk-ish (citizenship: 26 %, country of birth: 28 %) or North African (citizenship: 22 %, country of birth: 15 %) back-grounds. However, when in contact with school authori-ties as parents or guardians, this reason mostly affects respondents with Turkish background: every second respondent from this target group who experienced an incident of discrimination in this area of life identified their citizenship as the main reason (50 %).
For immigrants and descendants of immigrants from South Asia and Asia, the country of birth is the second most important reason for discrimination in healthcare (24 %) – after skin colour. Respondents with South Asian background in Greece appear particularly affected by this reason in access to healthcare (in addition to skin colour, mentioned by 72 %). Specifically, more than two thirds (73 %) of those who used healthcare services in the 12 months before the survey indicated their country of birth as the main reason for experiencing discrimina-tion (60 % mentioned their citizenship).
Citizenship and country of birth are mentioned more often by first-generation respondents than second-generation respondents in all areas.
Respondents’ accent or the way they speak the lan-guage of the survey country is considered as the main reason for discrimination by every fifth respondent in each of the following three areas of life: when using healthcare services (20 %), when looking for work (18 %), and when in contact with their children’s school as parents or guardians (19 %). In access to healthcare, the accent or the way one speaks the language of the survey country is mentioned as a reason for discrimina-tion mostly by Roma respondents (28 %) and by one out of four respondents with Turkish background who used healthcare services in the 12 months preceding the survey (25 %).
This specific reason is also mentioned more often by first-generation respondents than by second-genera-tion respondents in all areas.
The way respondents are dressed is equally important when looking for work (14 %) and in access to healthcare (13 %). One out of 10 respondents who felt discriminated against at work (10 %) identified the way they were
dressed as the main reason for the most recent incident. When it comes to this reason, the biggest differences between women and men are observed, with women mentioning it significantly more often than men.
2�1�2� Reporting discrimination
EU-MIDIS II asked respondents who experienced dis-crimination whether they reported or made a complaint about the most recent incident to anyone. Response options included reporting to a list of institutions, such as designated bodies or the police, as well as the place at which the incident occurred, such as the work place.
Overall, only one out of eight respondents (12%) reported or made a complaint about the most recent incident of discrimination based on ethnic or immigrant background (Figure 13).27 This means that the report-ing rate has not substantially changed since the first EU-MIDIS survey in 2008, when the overall reporting rate was 18 %. The low reporting level is even more problematic in light of the overall level of discrimination indicated in EU-MIDIS II (12 months before the survey: 24 %; five years before the survey: 38 %). These find-ings show that, despite efforts by the EU and its Member States, incidents of discrimination remain largely unre-ported and therefore invisible to institutions that have a legal obligation to respond to discrimination complaints.
Reporting rates among different groups and countries
The level of reporting discrimination is quite unevenly distributed across target groups and between the Mem-ber States and target groups, ranging from 30 % to 2 % (Figure 13). Looking at the results at the level of aggregate target groups, victims of discrimination with Sub-Saharan African and (South) Asian backgrounds tend to report more often than the average victim of discrimination (16% and 14%, respectively). By contrast, victims of discrimination who are recent immigrants or have a North African background tend to report less often than the average (9 % and 10 %, respectively).
As the following discussion shows, however, reporting levels substantially differ within the same target groups across the countries in which they were surveyed, and across different target groups within individual Mem-ber States, as well. For example, respondents who feel discriminated against on average tend to report more often in Finland, Ireland, the Netherlands and Sweden.
27 The reporting rate is calculated based on all respondents who experienced any discrimination in at least one of the areas of daily life covered in the survey. It would on average increase by around 2 % if the calculation were to consider only those respondents who indicated the exact timing of the latest incident they experienced.
43
What do the results show?
Figure 13: Respondents who reported or filed a complaint about last incident of discrimination, by target group and country (%) a,b,c,d
12
(3)
12(5)567
1114
151818
9
(10)
(5)(11)
(15)16
10(7)89
1829
12
16
1115
212122
168
(9)(9)
1112
151515
1725
2730
14
(8)
(3)(13)
(2)
0 402010 605030
LT
ES
EL
Group Average
SE
EL
MT
FR
FI
EE
Group Average
HU
HR
DK
Group Average
IT
LU
IE
LV
ES
NL
SK
NL
PT
DE
SE
Group Average
CY
PT
IT
CZ
AT
SI
AT
DK
UK
Total
IT
BE
RO
DE
Group Average
UK
Group Average
FR
BG
BE
PL
Group Average
TUR
NOAFR
(S)ASIA
RIMGR
ROMA
RUSMIN
EU-28
SSAFR
Notes: a Out of all respondents who experienced discrimination based on ethnic or immigrant background in at least one of the domains of daily life asked about in the survey (n=11,312, of which SSAFR: n=2,983, TUR: n=1,532, NOAFR: n=1,800, (S)ASIA: n=862, RIMGR: n=235, ROMA: n=3,730 and RUSMIN: n=170); weighted results.
b Results based on a small number of responses are statistically less reliable. Thus, results based on 20 to 49 unweighted observations in a group total or based on cells with fewer than 20 unweighted observations are noted in parentheses. Results based on fewer than 20 unweighted observations in a group total are not published.
c Acronyms for target groups refer to immigrants from [country/region] and descendants of immigrants from [country/region]: TUR = Turkey, SSAFR = Sub-Saharan Africa, NOAFR = North Africa, (S)ASIA = South Asia and Asia, RIMGR = recent immigrants from non-EU countries, RUSMIN = Russian minority, ROMA = Roma minority.
d Question: “Last time you felt discriminated against because of your ethnic or immigrant background at [domain], did you report or make a complaint about the incident?”
Source: FRA, EU-MIDIS II 2016
Second European Union Minorities and Discrimination Survey – Main results
44
Meanwhile, respondents in Belgium, Greece, Hungary, Poland, Portugal, Slovenia and Spain tend to report less often than the average. This indicates that rights con-sciousness – including the knowledge and means to complain – varies not only between individual respond-ents and/or target groups. It also points to varying degrees of effectiveness of existing laws and policies that aim to counteract discrimination and ensure equal-ity for all in the Member States. Moreover, it shows to which extent national equality bodies are successful in their efforts to reach out to the specific ethnic and immigrant communities surveyed in EU-MIDIS II.
SSAFR The highest reporting rate overall is observed for respondents with Sub-Saharan African background in Finland, where almost every third respondent who experienced discrimination reported or made a com-plaint about the latest incident (30 %). In Austria, Italy, and Portugal, not even one out of 10 respondents with Sub-Saharan African background did so (8 %, 9 % and 9 %, respectively). Meanwhile, respondents with Sub-Saharan African background in Ireland (27 %) and Sweden (25 %) indicate significantly higher levels of reporting than victims from this target group in Malta (11 %), Denmark (12 %), France and Germany (both 15 %).
TUR A look at reporting levels in the countries in which immigrants and descendants of immigrants from Tur-key were surveyed reveals a slightly lower variation in reporting rates than among other target groups. While one out of four victims from this group reported or made a complaint about the latest incident of dis-crimination in Sweden (22 %), Denmark (21 %), and the Netherlands (21 %), only every tenth did so in Germany (11 %). The rates in Austria and Belgium are similar, at 16 % and 15 %, respectively.
NOAFR A look at the results for respondents with North African background – the target group that on average indicated the highest discrimination rates – shows that reporting levels also differ widely across the countries in which this group was surveyed. While every third victim with North African background reported the latest incident of discrimination in the Netherlands (29 %), fewer than every tenth reported it in France (9%) and Belgium (8%). Almost none of the incidents experienced by respondents of the same target group were reported in Spain. Reporting rates were also low-est in Spain in EU-MIDIS I. At 18 %, the reporting rate for Italy is midway between the two extremes; compared to EU-MIDIS I, the rate for this particular group in this particular Member State has only slightly increased.
The relatively ‘high’ reporting rates in the Netherlands for both target groups surveyed there – respondents with North African and Turkish backgrounds – point to an increasing rights consciousness and, indirectly, to the effectiveness of the country’s institutions and
specialised bodies in reaching out to ethnic and immi-grant communities. In addition, the data on citizenship show that immigrants and descendants of immigrants in the Netherlands have among the highest rates of national citizenship, which might also explain this result.
ROMA In most Member States in which they were surveyed, Roma respondents continue to experience higher levels of discrimination than the average expe-rienced by all groups in those countries. They also continue to report the most recently experienced inci-dent less often than the average victims of discrimi-nation do. While nearly every fourth Roma victim of discrimination reported the latest incident in Croatia and Slovakia (both 18 %), just about one out of twenty victims of discrimination reported the most recent inci-dent in Greece (7 %), Hungary (6 %) and Spain (5 %). The findings also show that not much has changed since the first EU-MIDIS survey in terms of reporting levels. For example, in EU-MIDIS I, the reporting rate for Roma in Slovakia was more or less at the same level (20 %).
No substantial differences in the levels of reporting can be observed between first- and second-generation respondents or between different age groups.
Gender differences in reporting discrimination
Overall, female respondents are slightly more likely to report discrimination than male respondents (14 % and 11 %, respectively) (see Figure 14). This is particularly the case for immigrants and descendants of immi-grants from Turkey, where there is a nine-percentage-point difference between men and women (women 17 %, men 8 %). Similarly, female respondents from the North African target group are on average slightly more likely to file a complaint (12 %) than men from the same group (9 %). Apart from Roma respond-ents – among whom slightly more men than women reported the most recent incident of discrimination (men 13 %, women 11 %) – no further differences between men and women can be observed at the level of aggregate target groups.
Further, in-depth disaggregation of the results is not possible due to the low reporting rates overall. However, a look at the Member State level reveals some rela-tively prominent gender differences. In some countries, women victims of discrimination report more often than men (Germany: male: 7 %, female: 17 %; Denmark: male: 15 %, female: 21 %; Finland: male: 26 %, female: 36 %; United Kingdom: male: 14 %, female: 19 %). In Romania (male: 14 %, female: 8 %) and the Czech Republic (male: 18 %, female: 13 %), men report more often.
The next section (Figure 15) outlines reporting levels by gender for each area of life about which respondents were asked regarding their experiences of discrimination.
45
What do the results show?
Most complaints relate to experiences at work
As Figure 15 illustrates, overall, most reports or com-plaints about the latest incident of discrimination involve experiences of discrimination at work (40 %), followed by reports related to discrimination at administrative offices or public services (22 %). Reporting incidents related to respondents’ educational institutions ranks third (13 %) followed by reporting about experiences of unfair treatment when looking for work (11 %).
That most reports are related to incidents of discrimi-nation at work corresponds to the fact that, over all areas of daily life, most incidents were reported to
employers (36 %) and to trade/labour unions or staff committees (13 %), followed by the police (17 %) – the latter mostly related to entering a night club or bar. Of all reports made or complaints filed, only 4 % were directed towards an equality body.
These results clearly show that EU-MIDIS II respondents mostly experience discrimination when they look for work (see section on main reasons for discrimination in different domains). However, the majority of discrimina-tion victims in this specific life domain did not turn to any institution or specialised body regarding the most recent incident of discrimination. As further discussed later in this report, results show that respondents often
Figure 14: Respondents who reported or filed a complaint about last incident of discrimination, by target group and gender (%) a,b,c,d
0 402010 605030
161516Women
MenGroup Average
SSAFR
128
17WomenMen
Group AverageTUR
109
12WomenMen
Group AverageNOAFR
141414Women
MenGroup Average
(S)ASIA
9(9)(9)Women
MenGroup Average
RIMGR
1213
11WomenMen
Group AverageROMA
(3)(8)
(1)WomenMen
Group AverageRUSMIN
1211
14WomenMen
AverageTotal
Notes: a Out of all respondents who experienced discrimination based on ethnic or immigrant background in at least one of the domains of daily life asked about in the survey (‘n=11,312, of which men: n=6,285 and women: n=5,026, and SSAFR: n=2,983, TUR: n=1,532, NOAFR: n=1,800, (S)ASIA: n=862, RIMGR: n=235, ROMA: n=3,730 and RUSMIN: n=170); weighted results.
b Results based on a small number of responses are statistically less reliable. Thus, results based on 20 to 49 unweighted observations in a group total or based on cells with fewer than 20 unweighted observations are noted in parentheses. Results based on fewer than 20 unweighted observations in a group total are not published.
c Acronyms for target groups refer to immigrants from [country/region] and descendants of immigrants from [country/region]: TUR = Turkey, SSAFR = Sub-Saharan Africa, NOAFR = North Africa, (S)ASIA = South Asia and Asia, RIMGR = recent immigrants from non-EU countries, RUSMIN = Russian minority, ROMA = Roma minority.
d Question: “Last time you felt discriminated against because of your ethnic or immigrant background at [domain], did you report or make a complaint about the incident?”
Source: FRA, EU-MIDIS II 2016
Second European Union Minorities and Discrimination Survey – Main results
46
do not know that specialised bodies with legal man-dates to respond to discrimination complaints exist in each Member State.
On average, male victims of discrimination are more likely to report incidents that happen at work than female victims (45 % and 35 %, respectively). How-ever, more women than men report incidents relating to using administrative offices and public services (25 % and 18 %, respectively), as well as incidents of discrimination in shops or incidents related to their children’s schools. Finally, incidents related to entering a night club/bar/restaurant, and to using
healthcare services, are more often reported by men (Figure 15).
Second-generation respondents notably more often report incidents related to their own education (26 % vs 6 %) or to entering a shop (12 % vs 6 %) than first-generation respondents do. Compared to second-generation respondents, first-generation respondents more often report incidents concern-ing their children’s school (10 % vs 4 %), access to housing (12 % vs 4 %) and looking for work (11 % vs 7 %). A look at different age groups reveals a noteworthy gap between 16-to-24-year-old respond-ents and 25-to-44-year-old respondents regarding
Figure 15: Domains of daily life where last incident of discrimination based on ethnic or immigrant background was reported, by gender (%) a,b
0 402010 605030
131213
Own education
1112
10Looking for work
4045
35At work
2218
25Administrative offices/public service
108
11Trying to or entering a shop
99
8Housing
85
11Child’s education
878
Public transport
810
6Night club/bar/restaurant
78
6Using healthcare services
Women Men Average
Notes: a Out of all respondents who reported the last incident of discrimination based on ethnic or immigrant background in at least one of the domains of daily life asked about in the survey (n=1,521, of which men: n=790 and women: n=731); weighted results, sorted by average rate per domain.
b Question: “Last time you felt discriminated against because of your ethnic or immigrant background at [domain], did you report or make a complaint about the incident?”
Source: FRA, EU-MIDIS II 2016
47
What do the results show?
the latest incident of discrimination encountered in education: 42 % of the younger group reported these, while only 4 % in the older age group did so – which is also linked to respondents’ average ages when attending educational institutions.
Due to low reporting rates overall, this report limits its analysis for different domains and target groups to four aggregate target groups – immigrants and descend-ants of immigrants from Sub-Saharan Africa, Turkey, and North Africa, and Roma respondents (Figure 16). Apart from Roma respondents, discrimination victims from the other three target groups most frequently reported discrimination incidents that occurred at work. Roma respondents mostly reported incidents related to public services, to public transport, to looking for work or to entering a shop. Reports about discrimina-tion when trying to enter a night club/bar/restaurant
were most frequently made by Roma respondents and respondents with Sub-Saharan African background. Discrimination linked to respondents’ education was mostly reported by immigrants and descendants of immigrants from Turkey.
Satisfaction with how complaint was handled
Respondents who reported or filed a complaint about the most recent incident of discrimination based on ethnic or immigrant background were asked about their level of satisfaction with the way their com-plaint was handled by the authorities. Although most reports were made in connection with the workplace, respondents who made complaints to employers or trade unions were on average not quite satisfied with the way their complaints were dealt with. On a scale from one to four – where one meant ‘very dissatisfied’
Figure 16: Reports or complaints made about last incident of discrimination because of skin colour, ethnic origin or religion in all domains of daily life within SSAFR, TUR, NOAFR and ROMA target groups, all respondents, weighted and sorted by overall reporting rates (%) a,b,c,d
SSAFRAt work
Administrative offices/public service
Own education
Looking for work
Trying to or entering a shop
Child’s education
Housing
Night club/bar/restaurant
Public transport
Using healthcare services
44
22
9
13
11
12
8
4
4020 806010 5030 700
4
12
TURAt work
Administrative offices/public service
Own education
Looking for work
Trying to or entering a shop
Child’s education
Housing
Night club/bar/restaurant
Public transport
Using healthcare services
47
24
17
9
7
(1)
6
(6)
4020 806010 5030 700
14
(3)
Second European Union Minorities and Discrimination Survey – Main results
48
and four meant ‘very satisfied’ – average satisfaction levels were 2.2 with respect to employers, and 2.5 with respect to trade unions. Similarly, respondents who reported incidents to the police – in relation to discrimination experienced in or while trying to enter a night club or a bar – were also not satisfied with the way the police responded (mean value=1.8). The average satisfaction level of individuals who reported incidents connected to public services to a municipal-ity equals 1.6 – a value that falls between being totally dissatisfied and somewhat dissatisfied.
Not effective and not worth doing: reasons for non-reporting
Respondents who did not report the most recent incident of discrimination were asked to give the reasons for not doing so (Table 1). Overall, the find-ings show that there are remarkable differences between the reasons for not reporting discrimination among the various areas of life as well as among the different target groups.
Figure 16 (continued)
NOAFRAt work
Administrative offices/public service
Own education
Looking for work
Trying to or entering a shop
Child’s education
Housing
Night club/bar/restaurant
Public transport
Using healthcare services
43
20
16
7
4
(4)
5
(7)
4020 806010 5030 700
(9)
(10)
ROMAAt work
Administrative offices/public service
Own education
Looking for work
Trying to or entering a shop
Child’s education
Housing
Night club/bar/restaurant
Public transport
Using healthcare services
10
27
6
20
22
20
23
15
4020 806010 5030 700
10
7
Notes: a Out of all respondents who reported the last incident of discrimination based on ethnic or immigrant background in at least one of the domains of daily life asked about in the survey (n=1,521); weighted results.
b Acronyms for target groups refer to immigrants from [country/region] and descendants of immigrants from [country/region]: TUR = Turkey, SSAFR = Sub-Saharan Africa, NOAFR = North Africa, (S)ASIA = South Asia and Asia, RIMGR = recent immigrants from non-EU countries, RUSMIN = Russian minority, ROMA = Roma minority.
c Results based on a small number of responses are statistically less reliable. Thus, results based on 20 to 49 unweighted observations in a group total or based on cells with fewer than 20 unweighted observations are noted in parentheses. Results based on fewer than 20 unweighted observations in a group total are not published.
d Question: “Last time you felt discriminated against because of your ethnic or immigrant background at [domain], did you report or make a complaint about the incident?”
Source: FRA, EU-MIDIS II 2016
49
What do the results show?
Across all life domains asked about, respondents who did not report discrimination incidents to any organisation or authority said they mostly did not do so because they thought nothing would happen or change as a result of reporting. For example, with respect to access to hous-ing, 42 % indicated this reason. For incidents involving entering a bar or night club, or relating to administrative offices or public services, 39 % indicated this reason. The second most common reason given for not report-ing with respect to the majority of the domains was that respondents classified the incidents as too trivial or not worth reporting. Another widely mentioned reason was that discrimination happens all the time, with every fourth respondent noting this reason when explaining why they did not file complaints about discrimination relating to entering a shop (28 %), using public transport (25 %) or entering a night club or bar (24 %). Every fifth respondent also mentioned this reason in connection with incidents concerning their education (22 %), work-place (19 %) or access to healthcare (17 %).
The impression that nothing would happen or change as a result of reporting a discrimination incident endured while looking for work is most prominent among Roma respondents (47 %) and respondents with South Asian and Asian background (46 %), followed by respondents with Turkish (40 %), Sub-Saharan African (37 %) and North African backgrounds (30 %). This makes it by far the most common reason for not filing complaints among all
groups and across all countries. It is an especially impor-tant reason for Roma respondents in Portugal (98 %), where this reason holds back nearly all such respondents from filing complaints when encountering discrimination while looking for work. Two out of three Roma respond-ents in Bulgaria (66 %) gave this reason for not reporting. In Spain, only every fourth Roma respondent mentioned it (25 %). Respondents with Turkish background mostly did not report incidents because they thought nothing would change in the Netherlands (64 %). In Germany, only 34 % of the respondents from this target group provided this reason. For respondents with Sub-Saharan African background, this was mostly a reason for not reporting in Austria (70 %), Ireland (56 %) and Germany (54 %). Respondents with North African background mostly mentioned this reason in the Netherlands.
Having no proof was mainly identified as a reason for not reporting by respondents with Turkish background (30 %); in the area of looking for work, this was their second most common reason.
That the incident was too trivial and not worth report-ing was the reason most mentioned by respondents with South Asian and Asian backgrounds who looked for work – almost half (44 %) gave this reason. By contrast, this was only the third most common reason mentioned by the other target groups.
Table 1: Most common reasons for not reporting an incident in different areas, top three per area and top six overall, sorted by overall mention (%)
Look
ing
for w
ork
At w
ork
Heal
thca
re
Hous
ing
Scho
ol a
utho
ritie
s
Own
educ
atio
n
Nigh
t clu
b/ba
r/re
stau
rant
Adm
inis
trat
ive
offi
ces
or p
ublic
ser
vice
s
Publ
ic tr
ansp
ort
In o
r ent
erin
g a
shop
Nothing would happen/change by reporting discrimination
36 31 33 42 27 38 39 39 37 34
Too trivial/not worth reporting it 22 27 26 23 19 44 32 24 37 31
It happens all the time 18 19 17 11 7 22 24 18 25 28
I had no proof 19 16 15 25 17 12 20 24 12 12
I didn’t want to create trouble 11 16 17 13 12 13 14 15 8 14
I was concerned about negative consequences
7 18 12 5 33 15 9 15 4 7
Note: n Most common reason for not reporting in area; n Second most common reason for not reporting in area; n Third most common reason for not reporting in area.
Source: FRA, EU-MIDIS II 2016
Second European Union Minorities and Discrimination Survey – Main results
50
Having no proof of discrimination in access to housing mostly keeps respondents with Sub-Saharan African, North African and Turkish backgrounds from reporting. One out of four respondents (25 %) who did not report the last discrimination incident did so because they had no proof, making it the second most common reason for not reporting in this domain, followed by the rationale that the incident seemed too trivial or not worth report-ing (23 %). Particularly respondents with Sub-Saharan African and North African backgrounds (both 28 %) indicated lacking proof for reporting incidents in the domain of housing, followed by respondents with Turk-ish background, of which one out of five who did not report the latest incident (20 %) stated that they chose not to do so because they had no proof. Respondents from this target group deemed the incident at issue to be too trivial or not worth reporting more than any other group (38 %). Almost half of the respondents with Turk-ish background (47 %) and Roma respondents (45 %) who did not report the latest incident in the area of housing believed that nothing would happen or change after reporting, making this the most common reason in this domain for these two target groups. Moreover, every fourth Roma respondent did not know where to turn in case of discrimination in the area of hous-ing (24 %) – the most prominent reason given for not reporting by Roma respondents in the Czech Republic.
In the context of healthcare, almost half of Roma respondents (47 %) and every third respondent with Turkish background (33 %) mentioned the reason that nothing would happen or change to explain not reporting encountered discrimination.
In the area of education, respondents with Turkish and North African backgrounds and Roma respondents were most concerned about negative consequences. Concern about negative consequences of report-ing discrimination in their children’s school was most prominent among respondents with Turkish (57 %) and North African backgrounds (41 %), followed by Roma respondents (24 %). For Roma, this was the second most mentioned reason in this area – after the reason that nothing would happen or change (43 %). Incidents related to the respondents’ education seemed too trivial and therefore not worth reporting mostly to respond-ents with North African background (48 %); to those with Turkish background (44 %); as well as to 40 % of respondents with Sub-Saharan African background.
Across all areas of life, among Roma respondents, a prominent reason given for not reporting discriminatory incidents was that they did not know how to make a complaint or where to report it.
2�1�3� Awareness of support organisations, equality bodies and laws addressing discrimination
Awareness of organisations offering support or advice
The survey examined respondents’ level of awareness of organisations that offer support and advice in case of discrimination by asking whether they recognise one or more of up to three preselected equality bodies (in Germany, of up to four bodies; see Table 9 in Annex III). In addition, the survey asked respondents about their awareness of any organisations in their country of resi-dence that offer support or advice to people who have been discriminated against for whatever reason.
On average, most EU-MIDIS II respondents (71 %) were not aware of any organisations that offer support or advice to discrimination victims in their country of resi-dence (Figure 17). This could explain the low reporting rates. However, results vary widely across aggregate target groups and countries – between 96 % of recent immigrants in Slovenia and 54 % of respondents with Sub-Saharan African background in Sweden being una-ware of such organisations. Among respondents with North African background in Spain, recent immigrants in Slovenia, and respondents with South Asian background in Italy, almost none knew of such a support service or organisation. By contrast, respondents with Sub-Saha-ran African background in Sweden (44 %), from the Russian minority in Lithuania (40 %), and respondents in France – both with North African (36 %) and Turk-ish backgrounds (35 %) – show the highest awareness levels of such organisations.
The findings, outlined in Figure 17, clearly point to a country-specific effect with regard to the overall level of awareness about organisations that offer support. Independent of their origin, respondents in Sweden, the Netherlands, and France show similar levels of awareness, which are on average higher than in other countries. Specifically, in Sweden, for immigrants and descendants of immigrants from Sub-Saharan Africa, 44 %, and from Turkey, 35 %; in France, for immigrants and descendants of immigrants from North Africa, 36 %, and from Sub-Saharan Africa, 33 %; and in the Nether-lands, for immigrants and descendants of immigrants from North Africa, 31 %, and from Turkey, 29 %. Mean-while, independent of their origin, most respondents in Greece, Italy and Spain do not know of any organisa-tions that offer support to victims of discrimination.
There are some notable gender differences in the overall awareness level. Fewer women than men are aware of any organisations in Croatia (16 % vs 28 %), Ireland (21 % vs 31 %) and Italy (7 % vs 14 %). In other countries, the
51
What do the results show?
Figure 17: Awareness among all respondents of organisations that offer support or advice to victims of discrimination (regardless of the grounds of discrimination), by target group (%) a,b,c,d,e
No Don’t knowYes
4020 806010 5030 90700
LT
ES
HU
Group Average
SE
IT
PT
FI
SE
LV
Group Average
RO
HR
NL
Group Average
LU
IE
FR
EE
PT
FR
CZ
DE
IT
DE
UK
Group Average
CY
EL
NL
ES
BE
PL
MT
DK
AT
Total
EL
IT
SK
AT
Group Average
UK
Group Average
BE
BG
DK
SI
Group Average
TUR
NOAFR
(S)ASIA
RIMGR
ROMA
RUSMIN
EU-28
SSAFR
100
33 66 (1)
7426 (1)26 71 (3)26 71 (3)25 73 (2)
30 68 2
21 75 (3)19 80 1
17 81 (1)
12 79 (9)14 82 (4)
28 70 (2)
44 54 (2)
35 64 (1)
27 71 (2)22 78 (1)
20 76 (4)16 83 (1)
26 72 2
29 70 (2)
3 96
33 44 2222 76 2
17 8316 81 316 80 4
9 88 2
888 48215 3
5840 (2)5635 9
8015 56528 7
15 82 3
21 74 522 74 4
8 91 (1)
12 86 (2)
11 79 10
27 68 (5)
8 87 (5)21 76 3
36 64 (1)31 69
29 68 (3)11 87 2
6 91 331 68 1
7127 2
Notes: a Out of all respondents (n=25,515); weighted results. b Acronyms for target groups refer to immigrants from [country/region] and descendants of immigrants from
[country/region]: TUR = Turkey, SSAFR = Sub-Saharan Africa, NOAFR = North Africa, (S)ASIA = South Asia and Asia, RIMGR = recent immigrants from non-EU countries, RUSMIN = Russian minority, ROMA = Roma minority.
c Results based on a small number of responses are statistically less reliable. Thus, results based on 20 to 49 unweighted observations in a group total or based on cells with fewer than 20 unweighted observations are noted in parentheses. Results based on fewer than 20 unweighted observations in a group total are not published.
d Question: “Do you know any organisations in [COUNTRY] that offer support or advice to people who have been discriminated against – for whatever reason?”
e Responses that were either “don’t know”, “refused” and “doesn’t understand the question” are accounted for in the “Don’t know” category.
Source: FRA, EU-MIDIS II 2016
Second European Union Minorities and Discrimination Survey – Main results
52
opposite applies – such as in the Netherlands (women: 34 %, men: 26 %) and Poland (women: 30 %, men: 23 %).
Awareness of at least one equality body
Overall, the majority (62 %) of EU-MIDIS II respondents are not aware of any equality body in their country – although results vary by target group and country (see Figure 19 and Table 9 in Annex III, which lists the equality bodies presented to respondents). As Figure 18 shows, the best known equality bodies are in Ireland (67 %), Denmark (65 %), and the United Kingdom (60 %), where more than half of the respondents are aware of at least one equality body. In other countries, the proportion of respondents who know the equality bodies is low – for example, in Spain (6 %), Malta (9 %), Slovenia (10 %), Luxembourg (12 %), and Italy (14 %).
On average, more men (41 %) than women (35 %) are aware of at least one equality body, but differences between males and females vary on the individual coun-try level (Figure 18). The difference between awareness levels among men and women is particularly promi-nent in Bulgaria (men: 47 %, women: 27 %) and the United Kingdom (men: 65 %, women: 56 %). By con-trast, in Austria, Belgium, Finland, the Netherlands and
Poland, notably more women than men know of at least one equality body.
When asked about anti-discrimination legislation in their countries of residence, respondents on average showed high awareness levels, although results differ consider-ably across target groups and countries – ranging from 81 % of all respondents in France knowing of such a law to 18 % of all respondents in Malta. On average, most respondents (67 %) know that discrimination based on skin colour, ethnic origin or religion is unlawful in the coun-try in which they live (Figure 20). 18 % of all respondents think that there is no such law, while 16 % do not know whether such legislation exists. Overall, more men (70 %) than women (64 %) are aware of anti-discrimination laws.
The highest awareness of anti-discrimination legislation is found among respondents with Turkish background in Sweden (82 %), with Sub-Saharan African background in the United Kingdom (87 %) and France (81 %), and with North African background in France (81 %) and the Netherlands (78 %). The lowest awareness levels are found among Roma respondents in Portugal (13 %), respondents with South Asian background in Italy (15 %), and respondents with Sub-Saharan African background in Malta (18 %) (Figure 20).
Figure 18: Knowledge among all respondents of at least one equality body, by country and gender (%) a,b,c,d
4020 806010 5030 90700 100
IE
HR
SE
BE
LT
PL
CZ
CY
LV
UK
6769
65
4550
42
474748
4846
51
5254
50
5242
61
5253
52
555555
5660
53
6065
56
6568
60DK
Women
Men
Average
53
What do the results show?
Figure 18 (continued)
4020 806010 5030 90700 100
EU-2838
4135
ES6
85
MT9
8
SI10
12(9)
LU12
1411
IT14
188
AT21
1725
RO23
2620
EE24
2823
SK27
2628
HU313131
PT33
3530
FR36
3933
BG37
4727
DE37
4035
EL40
4435
FI42
3945
NL43
4047
Women
Men
Average
Notes: a Out of all respondents (n=25,515); weighted results. b Acronyms for target groups refer to immigrants from [country/region] and descendants of immigrants from
[country/region]: TUR = Turkey, SSAFR = Sub-Saharan Africa, NOAFR = North Africa, (S)ASIA = South Asia and Asia, RIMGR = recent immigrants from non-EU countries, RUSMIN = Russian minority, ROMA = Roma minority.
c Results based on a small number of responses are statistically less reliable. Thus, results based on 20 to 49 unweighted observations in a group total or based on cells with fewer than 20 unweighted observations are noted in parentheses. Results based on fewer than 20 unweighted observations in a group total are not published.
d Question: “Have you ever heard of the [NAME OF EQUALITY BODY]?”Source: FRA, EU-MIDIS II 2016
Second European Union Minorities and Discrimination Survey – Main results
54
Figure 19: Knowledge among all respondents of at least one equality body, by target group and country (%) a,b,c,d
4020 806010 5030 90700
LV
ES
SK
Group Average
DK
IT
AT
FI
IE
LT
Group Average
RO
CZ
SE
Group Average
IT
PT
UK
EE
PT
BE
HR
NL
LU
FR
DK
Group Average
CY
ES
NL
EL
BE
PL
MT
DE
SE
Total
EL
IT
HU
AT
Group Average
UK
Group Average
FR
BG
DE
SI
Group Average
TUR
NOAFR
(S)ASIA
RIMGR
ROMA
RUSMIN
EU-28
SSAFR
100
5
6765
6248
42
3432
2019
129
46
6645
4239
3821
38
52
5553
5210
5245
3937
3127
2317
1029
5652
2444
38
25
44
4810
4437
13
34
39
Notes: a Out of all respondents (n=25,515); weighted results. b Acronyms for target groups refer to immigrants from [country/region] and descendants of immigrants from
[country/region]: TUR = Turkey, SSAFR = Sub-Saharan Africa, NOAFR = North Africa, (S)ASIA = South Asia and Asia, RIMGR = recent immigrants from non-EU countries, RUSMIN = Russian minority, ROMA = Roma minority.
c Results based on a small number of responses are statistically less reliable. Thus, results based on 20 to 49 unweighted observations in a group total or based on cells with fewer than 20 unweighted observations are noted in parentheses. Results based on fewer than 20 unweighted observations in a group total are not published.
d Question: “Have you ever heard of the [NAME OF EQUALITY BODY]?”Source: FRA, EU-MIDIS II 2016
55
What do the results show?
Figure 20: Awareness among all respondents of laws prohibiting discrimination based on skin colour, ethnic origin or religion, by target group and Member State (%) a,b,c,d,e
4020 806010 5030 90700
LT
ES
HU
Group Average
SE
IT
AT
DE
UK
LV
Group Average
BG
CZ
NL
Group Average
PT
FI
FR
EE
ES
FR
HR
DE
IT
IE
DK
Group Average
UK
PT
NL
SK
BE
PL
MT
LU
SE
Total
EL
IT
EL
AT
Group Average
CY
Group Average
BE
RO
DK
SI
Group Average
TUR
NOAFR
(S)ASIA
RIMGR
ROMA
RUSMIN
EU-28
SSAFR
100
81 12 7
1276 1274 17 9
69 13 1864 14 21
79 12 9
55 26 1942 37 21
39 15 46
18 40 4227 42 31
77 13 10
87 7 6
82 11 7
65 17 1857 22 22
55 31 1342 33 24
64 18 18
71 18 11
38 24
74 17 947 8 44
51 24 2532 35 3331 28 40
28 35 38
2213 653536 29
1469 171851 31
1241 471550 35
31 41 29
54 32 1455 20 25
21 62 17
48 19 33
32 15 53
65 11 2437
15 48 3766 20 14
81 11 878 13 9
75 9 1533 32 35
30 45 2574 15 11
1867 16
No, there is no law Don’t knowYes, there is a lawNotes: a Out of all respondents (n=25,515); weighted results. b Acronyms for target groups refer to immigrants from [country/region] and descendants of immigrants from
[country/region]: TUR = Turkey, SSAFR = Sub-Saharan Africa, NOAFR = North Africa, (S)ASIA = South Asia and Asia, RIMGR = recent immigrants from non-EU countries, RUSMIN = Russian minority, ROMA = Roma minority.
c Results based on a small number of responses are statistically less reliable. Thus, results based on 20 to 49 unweighted observations in a group total or based on cells with fewer than 20 unweighted observations are noted in parentheses. Results based on fewer than 20 unweighted observations in a group total are not published.
d Question: “As far as you are aware, is there a law in [COUNTRY] that forbids discrimination based on skin colour, ethnic origin or religion?”
e Responses that were either “don’t know”, “refused” and “doesn’t understand the question” are accounted for in the “Don’t know” category.
Source: FRA, EU-MIDIS II 2016
Second European Union Minorities and Discrimination Survey – Main results
56
2�2� Hate crime – harassment and violence
KEY FINDINGS
n One in four respondents (24 %) experienced one or more incidents of hate-motivated harassment in the 12 months before the survey – that is, incidents which they perceived to be motivated by their ethnic or immi-grant background. Some 3 % experienced a hate-motivated physical attack in the year preceding the survey.
n Among the groups interviewed, Roma experienced the highest rate (30 %) of hate-motivated harassment in the 12 months before the survey, followed by immigrants and descendants of immigrants from North Af-rica (29 %). When compared with the 2008 EU-MIDIS survey, the latest data suggest an increase in harassment experienced by persons of North African background.
n Overall, second-generation respondents (32 %) report experiencing higher rates of hate-motivated harass-ment during the 12 months before the survey than first-generation respondents (21 %) do, as do young re-spondents (respondents aged 16-24 years: 32 %; 25-44 years: 27 %; 45-59 years: 19 %; 60+ years: 14 %).
n Considering both the prevalence and frequency of incidents, second-generation respondents indicate being doubly vulnerable to hate-motivated harassment – meaning that the overall prevalence of incidents is higher for the second generation than the first, and that the second generation is also more likely to experience recur-rent incidents. Specifically, 50 % of second-generation victims of hate-motivated harassment experienced six or more such incidents in the 12 months before the survey.
n In most cases (81 %), respondents felt that one or more of the incidents of harassment that they had experi-enced in the five years before the survey was motivated by their ethnic or immigrant background as opposed to any other reason, making these hate-motivated incidents. As for physical attacks, 57 % indicated that at least one of the incidents that they experienced in the five years before the survey was motivated by their ethnic or immigrant background.
n Most of the respondents’ experiences with hate-motivated harassment involved incidents in which the victim and offender came in direct contact with one another – for example, in the street or another location. The sur-vey results suggest that cyber-harassment involving personal insults or threats against immigrants and ethnic minorities is much less common than incidents that take place in person.
n Overall, the majority of hate-motivated incidents of harassment (71 %) and violence (64 %) were perpetrated by persons who were perceived to have no ethnic minority background.
n Victims often describe the perpetrators of hate-motivated harassment and violence as someone they did not know (72 % and 49 %, respectively). The next most often mentioned perpetrators are persons encountered at work, college or university, or in a training context (harassment – 15 %, violence 12 %), followed by neighbours (harassment – 8 %, violence – 13 %). Compared with men, incidents of hate-motivated violence against women were more often perpetrated by someone the women knew, such as a neighbour.
n Some 10 % of victims of harassment say that they reported the most recent incident – either to the police or to another organisation/service. Conversely, some 90 % of incidents are not brought to the attention of the police or other authorities and services, or the respondents are not sure whether the incident was ever brought to the attention of any authority or service.
n Hate-motivated violence is reported more often than harassment – 28 % reported the most recent incident of hate-motivated violence in the five years before the survey to the police or other organisation or service. Violent incidents are most often reported to the police, while harassment is usually reported to other organisa-tions or services – for example, to an individual at the place where the harassment happened.
n When asked why they decided not to report incidents of hate-motivated violence, respondents most often mention not being convinced that anything would happen or change as a result of reporting (41 %). Other common reasons for not reporting include that they preferred to deal with the problem themselves or with the help of family and friends (21 %), and that they perceived the incident to be minor and so not worth reporting (16 %). Furthermore, 11 % mentioned not trusting or being afraid of the police.
57
What do the results show?
Hate crime can affect anyone in society, and has impli-cations not only for the individual targeted, but for their family and their communities, and society as a whole, as well. It is the most severe expression of discrimination and a core fundamental rights abuse.
According to the Racial Equality Directive,28 harassment is a form of discrimination and is defined as “unwanted conduct related to racial or ethnic origin [that] takes place with the purpose or effect of violating the dig-nity of a person and of creating an intimidating, hos-tile, degrading, humiliating or offensive environment” (Article 2). Council Framework Decision 2008/913/JHA of 28 November 2008 on combating certain forms and expressions of racism and xenophobia by means of criminal law requires Member States to ensure that racist and xenophobic motivation is considered an aggravating circumstance or, alternatively, that such motivation may be taken into consideration by the courts when determining penalties. The 2012 Victims’ Rights Directive requires that “victims who have suf-fered a crime committed with a bias or discriminatory motive” receive an individual assessment to identify specific protection needs they may have (Article 22).29 The assessments prescribed in the Victims’ Rights Directive must take into account a person’s ethnicity, race and religion.
In 2016, the European Commission set up a High Level Group on combating racism, xenophobia and other forms of intolerance. A key objective of the high level group is to help ensure that hatred and intolerance are tackled more effectively. Within the framework of the high level group, FRA coordinates the work of a sub-group to assist Member States in developing effective methods for recording and collecting hate crime data. Current EU legislation, including the Council Framework Decision on Racism and Xenophobia,30 does not contain provisions for collecting and publishing data on hate crimes. Official data on incidents of criminal victimisa-tion carried out with a bias motive are not available for all EU Member States.31
In 2012, FRA published a report on Making hate crime visible in the European Union: acknowledging victims’
28 Council Directive 2000/43/EC of 29 June 2000 implementing the principle of equal treatment between persons irrespective of racial or ethnic origin, OJ L 180, 19 July 2000.
29 Directive 2012/29/EU of the European Parliament and of the Council of 25 October 2012 establishing minimum standards on the rights, support and protection of victims of crime, and replacing Council Framework Decision 2001/220/JHA, OJ L 315, 14 November 2012.
30 Council of the European Union (2008).31 See Chapter 3 on Racism, xenophobia and related intolerance
in FRA (2017a).
rights,32 in which the agency stressed the need to broaden the scope of hate crime data collection to make such crime visible in the EU, given that only few EU Member States collect and publish data covering a broad range of bias motivations. The report also noted European Court of Human Rights rulings expressing the need for states to ‘unmask’ the motivation behind racist offences as well as crimes committed because of the victim’s religious beliefs.
As part of the 2030 agenda for sustainable develop-ment, the United Nations has adopted the so-called Sustainable Development Goals (SDG) as well as related indicators, based on which progress in reaching these goals can be assessed. In connection with these indica-tors, all UN member states are to collect data concerning physical and sexual violence, as well as harassment. These and other SDG indicators are to be disaggregated, where relevant, by factors such as race, ethnicity and migratory status. The SDG indicators require data that can only be obtained through population-based sur-veys. The United Nations indicator database currently does not contain any data that would help assess pro-gress in terms of these indicators. For more details concerning EU-MIDIS II and the SDGs, see the section on the SDGs.
In addition to EU-MIDIS surveys, FRA has collected data on hate-motivated harassment and violence through its surveys on lesbian, gay, bisexual and trans (LGBT) persons, as well as on Jewish people. It has also carried out a dedicated survey on the experiences of Roma. The agency’s survey on violence against women asked similar questions concerning women’s experiences of physical violence and harassment in the 28 EU Mem-ber States. These surveys help provide an overview of hate-motivated harassment and violence. They have also produced consistent findings concerning the nature of incidents, consequence for victims, and obstacles – such as the failure to report violence and harassment to authorities – permitting better follow-up and service provision.
2�2�1� Experiences of harassment motivated by hatred
This section focuses on experiences with hate-moti-vated harassment – that is, experiences perceived by respondents as being connected to their ‘ethnic or immigrant background’. To put the results into context, respondents were also asked whether they had experi-enced harassment for any reason other than their ethnic
32 FRA (2012).
Second European Union Minorities and Discrimination Survey – Main results
58
or immigrant background. To reflect various situations where harassment can take place – both online and in person – respondents were asked about five forms of harassment: offensive or threatening comments in person; threats of violence in person; offensive ges-tures or inappropriate staring; offensive or threatening e-mails or text messages (SMS); and offensive com-ments made about them online. This means that, to qualify as harassment, the incident had to involve action that the respondent found ‘offensive’ or ‘threatening’, as opposed to actions that could be considered a normal part of everyday life. Victims of hate-motivated har-assment were given the opportunity to provide details concerning the most recent incident they experienced in the five years before the survey, including information about the perpetrator and whether the incident was reported to authorities or other services.
Prevalence and frequency of harassment motivated by hatred
One in four EU-MIDIS II respondents (24 %) experienced one or more incidents of harassment due to their ethnic or immigrant backgrounds in the 12 months before the survey (Figure 21). Roma experienced the highest rate of harassment: 30 % of Roma respondents indicated that they experienced harassment because they are Roma in the 12 months before the survey. Among immi-grants and descendants of immigrants, respondents with North African backgrounds indicate the highest rate of harassment (across five countries) – 29 % in the 12 months before the survey. At the other end of the scale, 6 % of Russian minority respondents indi-cated that they experienced harassment because of their Russian minority background in the 12 months before the survey.
Examining the experiences of various ethnic and immi-grant groups in more detail, the highest 12-month rate of hate-motivated harassment was recorded among Roma in the Czech Republic (56 %), followed by Roma in Greece (50 %), immigrants and descendants of immi-grants from Sub-Saharan Africa in Finland (47 %), immi-grants and descendants of immigrants from South Asia in Greece (41 %), and immigrants and descendants of immigrants from North Africa in the Netherlands (40 %). Although results concerning hate-motivated harassment are not fully comparable between EU-MIDIS I (2008) and EU-MIDIS II (2016) due to differences in the way the questions were formulated, in the 2008 survey, Roma in the Czech Republic and Greece also indicated the high-est rates of harassment out of all groups surveyed. On the other hand, while immigrants and descendants of immigrants from North Africa indicate the highest prev-alence of hate-motivated harassment among immigrant
groups in the 2016 survey, in the 2008 survey, this was highest among Sub-Saharan Africans. Therefore, it is possible that developments between the two surveys have made immigrants and descendants of immigrants from North Africa more vulnerable to hate-motivated harassment – or that the questions used in 2016 par-ticularly effectively capture the negative day-to-day experiences of this group.
In EU Member States where the survey interviewed more than one group, the different groups’ experiences within a country are generally aligned with each other in terms of the prevalence of hate-motivated harassment. However, some exceptions are worth noting. While 16 % of immigrants and descendants of immigrants from Turkey experienced hate-motivated harassment in Belgium in the 12 months before the survey, 38 % of immigrants and descendants of immigrants from North Africa experienced similar incidents in the coun-try. In Denmark, Germany and Sweden, immigrants and descendants of immigrants from Sub-Saharan Africa indicated experiencing hate-motivated harassment 10-13 percentage points more often than immigrants and descendants of immigrants from Turkey inter-viewed in these three countries.
The results differ greatly between countries and between groups. As noted, Roma in the Czech Repub-lic and Greece indicate the highest 12-month hate-motivated harassment rate in the survey (56 % and 50 %, respectively) – while rates are lower for Roma in Bulgaria (12 %), Hungary (18 %) and Portugal (20 %).
There is also a large difference in the 12-month rates of hate-motivated harassment experienced by immi-grants and descendants of immigrants from Sub-Saha-ran Africa. While some of the highest rates were found for this group in Finland (47 %), Ireland (38 %) and Luxembourg (38 %), the rates were lower in the United Kingdom (13 %), Malta (14 %) and Portugal (15 %). These differences may be due to differences in the demo-graphic composition of the groups and differences in immigration patterns, as well as relevant policies and measures implemented over the past years.
There are no gender differences in the reported rate of harassment based on respondents’ ethnic or immi-grant backgrounds in the 12 months before the survey – the rate is 24 % for both men and women. On the other hand, the survey shows that the rate of harass-ment experiences is higher among second-generation respondents (32 %) than first-generation respond-ents (21 %). These differences may be influenced by the different ways people of varying ages use public spaces and meet others – for example, young people
59
What do the results show?
Figure 21: Prevalence of harassment due to ethnic or immigrant background in 12 months before the survey (%) a, b, c
4020 806010 5030 90700
EE
ES
RO
Group Average
NL
UK
FR
DK
FI
LT
Group Average
PT
CZ
AT
Group Average
PT
AT
IE
LV
HU
NL
EL
DK
MT
SE
LU
Group Average
EL
BG
BE
SK
DE
PL
UK
IT
DE
Total
CY
FR
ES
BE
Group Average
IT
Group Average
IT
HR
SE
SI
Group Average
TUR
NOAFR
(S)ASIA
RIMGR
ROMA
RUSMIN
EU-28
SSAFR
100
24
473838
3533
3230
2415
1413
21
3732
2322
1916
23
40
4129
147
5650
3731
3027
2018
1230
86
46
24
10
23
1513
3833
29
29
33
Notes: a Out of all respondents (n=25,515); weighted results. b Acronyms for target groups refer to immigrants from [country/region] and their descendants: TUR = Turkey,
SSAFR = Sub-Saharan Africa, NOAFR = North Africa, (S)ASIA = South Asia and Asia, RIMGR = recent immigrants from non-EU countries, RUSMIN = Russian minority, ROMA = Roma minority.
C Question: “How many times have such incidents [that is, each of the five acts of harassment asked about in the survey] related to your ethnic or immigrant background happened in the past 12 months?”
Source: FRA, EU-MIDIS II 2016
Second European Union Minorities and Discrimination Survey – Main results
60
going out more than older persons. Indeed, respond-ents in the youngest age group note the highest rate of harassment based on ethnic or immigrant background (in the 12 months before the survey), and the preva-lence decreases with age: aged 16-24 years: 32 %; aged 25-44 years: 27 %; aged 45-59 years: 19 %; and aged 60+ years: 14 %.
Regarding the frequency of incidents, of those har-assed in the 12 months before the survey because of their ethnic or immigrant background, 20 % expe-rienced one incident, 37 % experienced 2-5 inci-dents, and 43 % experienced 6 or more incidents (including harassment taking place ‘all the time’). Second-generation respondents are more likely to experience harassment on a frequent basis than the first generation. Among the victims of harassment due to ethnic or immigrant background in the 12 months before the survey, 39 % of first-generation victims experienced 6 or more incidents (including ‘all the time’), while 50 % of second-generation vic-tims did so. Therefore, considering both prevalence and frequency (incidents), not only does a higher number of second-generation respondents experi-ence hate-motivated harassment, but they are also more likely to experience recurring incidents than the first generation.
There are no notable differences between the differ-ent age groups in terms of frequently experiencing harassment incidents. Among those who experi-enced harassment in the 12 months before the survey, between 40 % to 45 % – depending on the age group – experienced 6 or more incidents during this time.
Respondents were also asked whether they always or sometimes wear traditional or religious clothing when out in public, which sheds light on respondents’ expe-riences with harassment in connection with wearing such clothing. Among the groups interviewed, this type of clothing was most common among Muslim respondents; their experiences with harassment – with particular attention to experiences perceived to be related to wearing traditional or religious clothing – are presented in a separate FRA report.33 That analysis showed that Muslim women who at least sometimes wear traditional or religious clothing when out in pub-lic particularly experience higher rates of harassment (31 % in the 12 months before the survey, compared with 23 % among Muslim women who do not wear such clothing in public places).
Respondents were also asked whether they expe-rienced, in the five years before the survey, har-assment for any reason, including in connection with another bias motivation – e.g. gender, age or
33 FRA (2017b).
disability – and without any bias motivation: 19 % said they experienced harassment.
Type of harassment experienced
The survey asked respondents specific questions about five types of harassment (including cyber-harassment): offensive or threatening comments in person; threats of violence in person; offensive gestures or inap-propriate staring; offensive or threatening e-mails or text messages (SMS); and offensive comments made about them online.
Most of the harassment experiences shared by respond-ents involve situations where the offender and the vic-tim find themselves in the same space – the in-person harassment rate in the 12 months before the survey was 24 %, while 3 % experienced cyber-harassment during this timeframe. The most common form of harassment consists of offensive gestures or inappro-priate staring (18 %), followed by offensive or threat-ening comments (16 %) (Figure 22). The other forms of harassment have prevalence rates of 4 % or less. Cyber-harassment is highest towards young respond-ents and decreases with age (respondents 16-24 years of age: 6 %; 25-44 years: 3 %; 45-59 years: 1 %; 60+ years: close to 0 %). Similar results concerning age and cyber-harassment were found in the agency’s survey on violence against women.34
The responses do not point to notable gender differ-ences. In-person and cyber-harassment rates in the 12 months before the survey were somewhat higher towards second-generation respondents (20 %) than first-generation respondents (31 %).
Most respondents experienced harassment motivated by hatred ‘offline’, from individuals they came across in person, rather than ‘online’ – a context in which they can be more selective in terms of who to contact.
Perpetrators of harassment motivated by hatred
Respondents who experienced harassment based on their ethnic or immigrant backgrounds were asked about the perpetrator(s)’ characteristics. For example, respondents could indicate whether the perpetrator was someone from work, a police officer or a border guard, or somebody who they clearly identified as a member of a right-wing extremist/racist group. These results can help identify the context in which hate-motivated harassment most commonly occurs, which in turn can support developing more targeted policy responses on combating harassment, encouraging reporting and assisting victims.
34 FRA (2014).
61
What do the results show?
Perpetrators of hate-motivated harassment – specifi-cally, of the most recent incident in the five years before the survey – were most often described as an unknown person: in 72 % of cases. This was followed by some-body at work/in college or university/training – 15 %, and neighbours – 8 %. Meanwhile, 3 % of respondents who experienced hate-motivated harassment iden-tified the perpetrator as a police officer or a border guard, 3 % said the perpetrator was a member of a right-wing extremist/racist group, and 4 % said it was a public official. These results are broadly in line with those of the previous EU-MIDIS survey, which showed that – across the groups interviewed – perpetrators of hate-motivated harassment were most often people whom the respondents did not know.
Notably, 21 % of Roma in Portugal and 10 % of Roma in Spain who experienced hate-motivated harassment due to their ethnic or immigrant background indicated that the perpetrator was a police officer or a border guard. These experiences contrast with the results for the other groups interviewed in these two countries. Among immigrants and descendants of immigrants from Sub-Saharan Africa (in Portugal) and North Africa (in Spain) very few had experienced hate-motivated harassment by police or border guards (2 % and 1 %, respectively – although these results should be inter-preted with caution due to the small number of cases available for analysis).
There is no notable difference in terms of who was identified as the perpetrator of the most recent inci-dence of hate-motivated harassment between inci-dents experienced by women and men. In the case of hate-motivated harassment incidents experienced by second-generation respondents, the perpetrators were slightly more often described as someone at work/in college or university/training (20 %) than by the first generation (13 %). This may partly be due to the age dif-ference between the first and second generation: sec-ond-generation respondents are on average younger and so may, compared to first-generation immigrants, have more recent experiences with incidents related to college, university or other forms of training. There are no notable differences between first- and second-generation respondents with respect to the other per-petrator categories for hate-motivated harassment.
Respondents in the oldest age group (60+ years) less often described the perpetrator as someone they did not know (63 %), when compared with all respondents (72 %). Over 60-year-old victims of hate-motivated harassment more often described the per-petrator as an acquaintance, friend or relative (12 %) than all respondents did (5 %). Respondents in the youngest age group (16-24 years) were most likely to indicate that the perpetrator was someone at work/in college or university/training (19 %), com-pared with 14 % among 25-44-year-olds and 13 % among 45-59-year-olds (for the oldest age group,
Figure 22: Types of hate-motivated harassment experienced in 12 months before the survey (%) a,b
0
40
20
60
10
50
30
Offensive gestures or inappropriate
staring
Offensive or threatening comments
Threatened with violence
in person
Offensive, personal comments
on the internet
Offensive emails or text
messages
1816
4 2 1
Notes: a Out of all respondents (n=25,515); weighted results. b Question: “How many times have such incidents [that is, each of the five acts of harassment asked about in
the survey] related to your ethnic or immigrant background happened in the past 12 months?”Source: FRA, EU-MIDIS II 2016
Second European Union Minorities and Discrimination Survey – Main results
62
60+ years, the number of respondents who experi-enced hate-motivated harassment in this context was too low for analysis).
The survey respondents were also asked about the perceived ethnic or immigrant backgrounds of per-petrators of hate-motivated harassment. This infor-mation helps assess to which extent hate-motivated harassment is carried out by persons perceived to have no discernible ethnic minority background in the country in which they live, or by persons with the same ethnic minority backgrounds as victims or other ethnic minority backgrounds. Overall, the majority of hate-motivated harassment incidents described in the survey (71 %) were perpetrated by a person who was perceived to have no ethnic minority background. Meanwhile, 23 % of victims of hate-motivated harassment said the perpetrator was from another ethnic minority, and 8 % said that the perpetrator had the same ethnic or immigrant back-ground as themselves. This corroborates the finding of the first EU-MIDIS survey, in which persons from the ‘majority’ population were more often identified as perpetrators of harassment than persons from ethnic minority groups.
Among the main groups surveyed in EU-MIDIS II, Roma are more likely than other groups to experi-ence incidents perpetrated by individuals who belong to the same group as themselves (17 %). This could be because some Roma live in areas where most other inhabitants are also Roma. Immigrants and descendants of immigrants from Sub-Saharan Africa were most likely to experience incidents of hate-motivated harassment perpetrated by individuals from another ethnic minority: 31 %, compared to the average of 23 % among all respondents. Immigrants and descendants of immigrants from Sub-Saharan Africa in France (35 %) and Sweden (44 %) were particularly likely to indicate that this was the case. Roma in Bulgaria (42 %) and Romania (40 %) as well as immigrants and descendants of immigrants from Asia living in Cyprus (45 %) also often indicated that the perpetrator had an ethnic minority background other than their own.
There are no notable differences between women and men and the background of the perpetrators of hate-motivated harassment, nor between first- and second-generation respondents. Similarly, differences between different age groups are small and not con-sistent. However, some differences can be discerned based on age: 26 % of 25 to 44 year-old respondents said that the perpetrator of hate-motivated harass-ment was from another ethnic minority group, com-pared with 18 % of 45 to 59-year-old respondents. This may reflect different patterns of daily life in different
age groups, bringing younger respondents into more frequent contact – and conflict – with persons from other minority groups than older respondents.
Reporting harassment motivated by hatred and reasons for not reporting
Overall, 10 % of harassment victims said that they reported the most recent incident – either to the police or to another organisation/service (Figure 23). Con-versely, some 90 % of incidents were not brought to the attention of the police or other authorities and services, or respondents were not sure whether or not the incident was ever brought to their attention. This means that a large majority of victims of harassment motivated by hatred may have missed out on their opportunity to learn about their rights in the situa-tion and about options available to them in terms of recourse and support. The results also suggest that rates of reporting harassment incidents to the police have not increased since the first EU-MIDIS survey was conducted in 2008.
Of those who reported the most recent incident of hate-motivated harassment in the five years before the survey, 36 % reported it to the police, 53 % to another organisation/service, and 10 % reported both to the police and to another organisation/ser-vice. Apart from the police, among other organisations and services contacted following an incident of hate-motivated harassment, victims most often mentioned contacting someone in the organisation/institution in which the incident took place (37 % of all those who reported an incident anywhere), or contacting an ‘other organisation’ (9 %). Of the 8,709 respond-ents who provided more details regarding the most recent incident of hate-motivated harassment they endured in the five years before the survey – and of the 708 who reported the incident – 13 respondents across the EU-28 said that they contacted a national equality body, human rights institution or ombudsman as a result of that incident.
There are no notable differences between women and men, or between first- and second-generation respondents, in terms of reporting the most recent inci-dent of harassment – neither concerning the reporting rate overall nor related to where the report was made.
The most common reason for not reporting the most recent harassment incident to the police or any other organisation was the perception that nothing would happen or change by reporting the incident (41 %). This was followed by the rationale that ‘the incident was minor and not worth reporting, it happens all the time’ (38 %); that reporting would be too bureaucratic or time-consuming (12 %); and ‘I dealt with the problem
63
What do the results show?
myself or with help from family or friends’ (11 %). Other reasons were mentioned by under 10 % of harassment victims (respondents could provide multiple reasons).
There are no notable differences in reasons given by women and men for not reporting the most recent harassment incident. Second-generation respond-ents who experienced hate-motivated harassment were more likely to feel that nothing would happen or change by reporting the incident (47 %) than first-generation respondents (38 %). Other than that, there are no notable differences in the reasons given by first- and second-generation respondents for not reporting incidents of harassment.
Among respondents who reported the most recent incident of hate-motivated harassment to the police, 63 % were dissatisfied with the way police dealt with the matter. Men were more often dissatisfied with the police’s response (67 %) than women (59 %). Second-generation respondents who reported the most recent incident to the police were particularly often dissatis-fied with the police response (87 %) compared with first-generation respondents (52 %). It is possible that second-generation respondents – most of whom have been educated in the country in which they live and are familiar with the national administration – have higher expectations regarding the police’s response than first-generation immigrants do.
2�2�2� Experiences of physical violence motivated by hatred
This section describes incidents which respondents perceived to have occurred due to their ethnic or immigrant background, and which involved somebody physically attacking them – for example, incidents where the perpetrator hit, pushed, kicked or grabbed the respondent. The survey also gave respondents an opportunity to provide information about physical attacks that may have occurred for other reasons, to place the hate-motivated incidents into context. In addition to the prevalence and frequency of these incidents, the survey asked about details of the inci-dents, such as characteristics of the perpetrators and whether the incidents were reported anywhere. Such context can be helpful for identifying opportunities to prevent hate-motivated violence as well as for ensur-ing that as many victims as possible are informed about their rights and available support.
Prevalence and frequency of violence motivated by hatred
In EU-MIDIS II, 3 % of all respondents indicated that they experienced hate-motivated violence – that is, one or more physical attacks – due to their ethnic or immigrant background in the 12 months before the survey. This suggests that the overall situation has remained static since the 2008 survey, in which 4 % of respondents indicated having experienced a physical
Figure 23: Reporting hate-motivated harassment to authorities or services – most recent incident in 5 years before the survey (%) a,b
10
89
1
YesNoDon’t know/No answer
Notes: a Out of all respondents who experienced hate-motivated harassment in the five years before the survey (n=8,709); weighted results.
b Question: “Thinking about this LAST incident [that is, hate-motivated harassment experienced in the five years before the survey], did you report it or make a complaint about it?”
Source: FRA, EU-MIDIS II 2016
Second European Union Minorities and Discrimination Survey – Main results
64
attack or threats of violence35 motivated by their immigrant or minority background in the 12 months before the survey.
To contextualise the results on hate-motivated incidents, respondents were also asked whether they had experi-enced a physical attack for any reason in the five years before the survey – that is, including hate-motivated inci-dents as well as any other incidents in which the respond-ents were physically attacked. Most of the respondents who experienced a physical attack (57 %) indicated that at least one of the incidents they experienced in the five years before the survey was motivated by their ethnic or immigrant background. On the other hand, 43 % of the victims had experienced a physical attack but did not perceive it as having to do with their ethnic or immigrant background; these incidents could have involved another bias motivation – such as being perpetrated against per-sons with a disability, or based on their perceived sexual orientation – but the survey did not ask questions that would allow analysing this in more detail.
The highest 12-month rate of hate-motivated violence was recorded for Roma in Slovakia (11 %), followed by immigrants and descendants of immigrants from Sub-Saharan Africa in Austria (10 %). Out of all countries and
35 The 2008 survey asked respondents whether they had been personally attacked or threatened in a way that really frightened them, whereas the question used in 2016 refers only to physical violence (and not threats).
groups, eight groups with the highest rate of experienced hate-motivated violence are either Roma or of Sub-Saha-ran African background (Figure 24). However, the survey results suggest that the experiences of Roma as well as of other groups vary greatly depending on the coun-try in which they live. While the 12-month rate of such violence against Roma and persons with Sub-Saharan African background is high in some countries – compared with all groups and countries surveyed – the victimisa-tion rates for these groups are among the lowest of all groups in EU-MIDIS II with respect to Roma in Bulgaria and Portugal (in both cases, close to 0 % victimised in the 12 months before the survey), and immigrants and descendants of immigrants from Sub-Saharan Africa in Portugal and the United Kingdom (in both cases 1 %).36
Overall, there are no notable differences in the preva-lence of hate-motivated violence in the 12 months before the survey for men (3 %) and women (2 %), or between first-generation (2 %) and second-generation respondents (3 %). While differences are also small with respect to different ages, the results on hate-motivated violence consistently show that the prevalence is high-est for the youngest age group and decreases with age: 16-24 years old – 4 %; 25-44 years – 3 %; 45-59 years – 2 %; and 60+ years – 1 %.
36 Due to the rare nature of hate-motivated violence against Roma in Bulgaria and Portugal, and against immigrants and descendants of immigrants from Sub-Saharan Africa in Portugal and the United Kingdom, the results are based on a small number of cases and may therefore be less reliable.
Figure 24: Ten groups with the highest rates of hate-motivated physical violence in 12 months before the survey, out of all groups surveyed (%) a,b,c,d
0 402010 605030
4555556
7
CZ - ROMAEL - ROMAEL - SASIA
AT - TUR
HR - ROMA
FI - SSAFR
DE - SSAFR
MT - SSAFR
1011SK - ROMA
AT - SSAFR
Notes: a Out of all respondents (n=25,515); weighted results. b For several groups which are not presented in this figure, the number of victims of hate-motivated physical assault
in the 12 months before the survey is very small, which limits the extent of the analysis. Therefore, the results presented here refer only to the groups with the highest rates.
c Acronyms for target groups refer to immigrants from [country/region] and their descendants: TUR = Turkey, SSAFR = Sub-Saharan Africa, NOAFR = North Africa, (S)ASIA = South Asia and Asia, RIMGR = recent immigrants from non-EU countries, RUSMIN = Russian minority, ROMA = Roma minority.
d Question: “How many times has this happened [that is, hate motivated physical attack] in the past 12 months because of your ethnic or immigrant background?”
Source: FRA, EU-MIDIS II 2016
65
What do the results show?
The prevalence of experienced hate-motivated vio-lence in the 12 months before the survey is the same for respondents who indicated wearing traditional or religious clothing and respondents who do not wear such clothing, including when the victims’ gender is taken into account: the 12-month prevalence is 3 % for men, and 2 % for women, irrespective of whether or not they wear traditional clothing in public. These results refer to the experiences of all respondents in the survey; FRA presented a more detailed analysis of experiences of Muslim respondents in a separate report, including on experiences of Muslims who at least sometimes wear traditional or religious clothing when out in public.37 Those results suggest that Mus-lims who wear such clothing are as likely to experience hate-motivated violence as Muslims who do not wear traditional or religious clothing (including headscarves or niqabs for women). By contrast, regarding Muslim women and men’s experiences with hate-motivated harassment, a higher rate of incidents was observed for women and men who at least sometimes wear traditional or religious clothing in public, than for Mus-lim women and men who do not wear such clothing.
Most respondents who experienced hate-motivated violence in the 12 months before the survey experi-enced one incident (58 %), while 33 % experienced two to five incidents and 9 % experienced six or more incidents. Experiencing frequent hate-motivated physical attacks – six or more incidents in a year – is more common among male respondents than female respondents. Some 11 % of men who experienced hate-motivated violence experienced six or more inci-dents in the 12 months before the survey, compared with 6 % of women. This specific result concerning women victims is based on a small number of cases and should be interpreted with caution.
Experiencing multiple incidents of hate-motivated violence in the 12 months before the survey is more common among second-generation respondents than first-generation respondents. Some 28 % of victims of hate-motivated violence who are first-generation immigrants experienced two to five incidents in the 12 months before the survey, compared with 38 % of victims among the second generation. The results for the category ‘6 or more times’ involve a low number of cases, which limits further analysis by immigrant generation. Taking respondents’ ages into consid-eration when analysing the frequency of hate-moti-vated violence is similarly hampered by low numbers in the survey.
37 FRA (2017b).
Perpetrators of violence motivated by hatred
When asked about the perpetrator of the most recent incident of hate-motivated violence they experienced in the five years before the survey, half of the victims (49 %) indicated that the perpetrator was someone they did not know beforehand. The next most often cited type of perpetrator was a neighbour (13 %) or somebody from work, college or training (12 %). Mean-while, 5 % described the perpetrator as a member of a right-wing extremist/racist group.
Notably, 10 % of victims indicated that the perpetrator was a police officer or a border guard. Due to small cell counts at the country level, it is not possible to assess in more detail the experiences of specific ethnic or immi-grant groups and their experiences of hate-motivated violence by police officers or border guards. Elsewhere in the survey, respondents were also separately asked whether they had been physically assaulted by a police officer in an incident that was perceived to have been motivated by their ethnic or immigrant background. Overall, 2 % of respondents had experienced such an incident in the five years before the survey, and 1 % had experienced this over the 12 months before the survey. Further examination of these results is severely limited by the small number of cases available for analysis. However, the data suggest that a majority of those assaulted are men – 3 % male respondents said that they had been assaulted by the police in the 5 years before the survey, compared with close to 0 % of women. The highest five-year prevalence of being physically assaulted by the police was found among immigrants and descendants of immigrants from South Asia in Greece (6 %), immigrants and descendants of immigrants from Sub-Saharan Africa in Austria (5 %), and Roma in Spain and Croatia (both 4 %).
In terms of hate-motivated violence overall, women were less likely to have experienced hate-motivated violence by an unknown person than men (women – 44 %, men – 51 %). Instead, women more often said that the perpetrator was a neighbour (women – 20 %, men – 9 %) or an acquaintance/friend/relative (women – 14 %, men – 6 %). Efforts to address hate-motivated violence should therefore recognise that women may be more at risk of experiencing incidents by known perpetra-tors in their close surroundings, while men more often experience incidents perpetrated by persons previously unknown to them.
The survey also asked respondents whether they would describe the perpetrator as having no ethnic minor-ity background in the country in which they live, or whether they perceived the perpetrator as having the same or a different ethnic or immigrant background as
Second European Union Minorities and Discrimination Survey – Main results
66
themselves. 64 % of victims described the perpetrator of the most recent incident of hate-motivated violence as someone who did not have an ethnic minority back-ground. For 13 %, the perpetrator had the same ethnic minority background as themselves, and 32 % said that the perpetrator had a different ethnic minority back-ground than their own. The sum of these percentages is over 100 % because some incidents involve multiple perpetrators with different ethnic backgrounds.
Due to the small number of cases involved, it is not pos-sible to assess differences in terms of the backgrounds of the perpetrator(s) at the country level. With respect to the main groups interviewed and their experiences in the five years before the survey, 35 % of Roma vic-tims of hate-motivated violence said that the perpetra-tor was also Roma, compared with the average of 13 % of respondents indicating overall that the perpetrator had the same ethnic or immigrant background as them-selves. For other groups interviewed, perpetrators from the same ethnic or immigrant group as respondents were not mentioned as frequently, and the results cannot be analysed further due to the small number of cases avail-able for analysis. As already noted earlier, the higher rate of intra-group incidents among Roma may partly be explained by the fact that some Roma respondents live in areas where most other people are also Roma.
For immigrants and descendant of immigrants from Asia and South Asia, the most recent incident of hate-motivated violence in the past five years was more often perpetrated by individuals from other minority backgrounds (48 %) compared with respondents over-all (32 %). Immigrants and descendants of immigrants from North Africa encountered the highest proportion of hate-motivated violent incidents perpetrated by mem-bers of the majority population (77 %). Close to half of Roma – 45 % –indicated that the perpetrator of the most recent incident of hate-motivated violence was some-one with no ethnic minority background, compared to 64 % of all respondents.
Women indicated more often than men that the perpe-trator had another ethnic minority background – 41 %, compared with 26 % of men. On the other hand, 19 % of women said that the perpetrator had the same ethnic background, compared with 9 % of men. Men were more likely to experience hate-motivated violence by someone with no ethnic minority background than women – 67 % for men, 59 % for women.
Analysis of the results concerning the backgrounds of the perpetrators and the immigrant generation of the victims is limited by the small number of cases in some categories. The largest difference between first- and second-generation respondents can be found in the percentage indicating that the perpetrator had another ethnic minority background – 26 % of first-generation
respondents (victims of hate-motivated violence) indi-cated this, compared with 37 % of second-generation respondents. In terms of victims’ age, older respondents more often indicated having experienced hate-moti-vated violence involving a perpetrator who did not have ethnic minority background (74 %) than young respond-ents (62 %). In other cases – concerning perpetrators who had the same background as the respondent or another minority background – the differences between age groups are small or not consistent in their direction.
Reporting violence motivated by hatred and reasons for not reporting
Of the victims of hate-motivated violence who reported the most recent incident in the 5 years before the sur-vey, 28 % reported it to the police or another organisa-tion or service. Specifically, 56 % contacted the police, 30 % another organisation/service, and 13 % reported both to the police and somewhere else. Among other organisations/services contacted following an incident of hate-motivated violence, victims most often men-tioned contacting someone in the organisation or insti-tution in which the incident took place (11 % of all those who reported an incident anywhere) or contacting an ‘other organisation’ (10 %). The rest of the respond-ents who reported hate-motivated violence mentioned a variety of organisations and services among those listed in the survey, such as victim support services, legal professionals, healthcare services, and national equality bodies, human rights institutions and ombuds-persons – each contacted by only few respondents. A comparison with results from the first EU-MIDIS survey in 2008 suggests that reporting incidents of violence has not increased between then and 2016, when the second survey was conducted.
There is no difference between reporting rates for men and women in terms of reporting the most recent incident of hate-motivated violence to police or any other organisation/service. First-generation immi-grants report incidents somewhat more often (32 %) than second-generation respondents (25 %) – possible reasons for this result can be identified in the reasons respondents gave in the survey for not reporting inci-dents, described later in this section. Individuals who fall into age groups in the middle of the scale report incidents more often than respondents in the youngest or oldest age groups (16-24 years – 19 %; 25-44 years – 31 %; 45-59 years – 40 %; 60+years – 22 %).
Respondents who did not report the most recent incident of hate-motivated violence encountered in the five years before the survey most often indicated that they were not convinced that anything would happen or change by reporting (41 %) (Figure 25). Other common reasons for not reporting included dealing with the problem oneself or with the help of family and friends (21 %) and the
67
What do the results show?
perception that the incident was minor and therefore not worth reporting (16 %). Furthermore, 11 % mentioned not trusting the police or being afraid of the police.
The reasons for not reporting an incident of hate-motivated violence were largely the same for women and men. Women are somewhat more often concerned about intimidation/retaliation by the perpetrator (men – 8 %, women – 18 %). Here it is worth remembering that women, more often than men, indicated that the perpetrator was someone they knew, such as a neigh-bour, acquaintance, friend or relative; it might be more difficult for women to avoid contact with these persons than it is when the perpetrator is an unknown person. Men, on the other hand, believe that reporting would be bureaucratic or time consuming (16 %) at a higher rate than women (5 %).
When asked why they did not report the most recent incident of hate-motivated violence, 30 % of second-generation respondents said that they dealt with the problem themselves or with the help of family and friends, compared with 11 % of first-generation immigrants. Compared with the first generation,
second-generation respondents also more often men-tioned that, in their opinion, the incident was minor (10 % among the first generation, 24 % of the second generation) or that nothing would happen or change by reporting (38 % among the first generation, 48 % of the second generation). By contrast, first-generation respondents more often mentioned not trusting the police or being afraid of the police (first generation – 14 %, second generation – 5 %).
2�2�3� Harassment and physical violence against respondents’ family or friends – incidents motivated by hatred
Across the EU, 25 % of respondents are aware of some-one in their circle of family or friends being insulted or called names because of their ethnic or immigrant back-ground in the 12 months before the survey. Among some of the groups, this rate is much higher, with half or more of respondents aware of such incidents, specifically: Roma in the Czech Republic (57 %); immigrants and
Figure 25: Reasons given for not reporting hate-motivated violence to police (%) a,b,c
0 402010 605030
1
1
5
7
11
11
12
12
16
21
41Nothing would happen or change by reporting the incident(s)
I dealt with the problem myself or with help from family or friends
The incident was minor and not worth reporting, it happens all the time
It would have been too bureaucratic, time-consuming
I was afraid of intimidation/ retaliation from perpetrators
I don’t trust the police/ I was afraid of the police
I was concerned that no one would believe me or take me seriously
Somebody stopped me or discouraged me
I did not know where to go/ whom to contact
I couldn’t report it because of a residence permit problem
Language barrier (could not report because of language difficulties)
Notes: a Out of all respondents who did not report the most recent incident of hate-motivated violence to the police in the five years before the survey (n=994); weighted results. Respondents whose answer was recorded as “Don’t know”, “Doesn’t understand the question” or “Refused” are included in the total but these answer categories received only few answers and are excluded from the figure.
b Question: “Why did you NOT report the incident or make a complaint to the police or any other organisation?” c Respondents were able to give more than one answer, so categories may total to more than 100 %.Source: FRA, EU-MIDIS II 2016
Second European Union Minorities and Discrimination Survey – Main results
68
descendants of immigrants from Turkey (51%) and North Africa (49 %) in the Netherlands; and Roma in Greece (49 %). The lowest rate was found among respondents from the Russian minority in the three Baltic States and recent immigrants in Slovenia – in each case under 10 %.
There is no notable difference in terms of women and men being aware of family members or friends being insulted or called names because of their ethnic or immigrant background in the 12 months before the sur-vey (women – 25 %, men – 26 %). Second-generation respondents more often know of a family member or friend who has encountered such behaviour in the 12 months before the survey than first-generation respond-ents (first generation – 22 %, second generation – 35 %).
Overall, 9 % of respondents know of someone in their circle of family or friends who has been physically attacked because of their ethnic or immigrant back-ground in the 12 months before the survey. While 34 % of Roma in the Czech Republic and 28 % of immigrants
and descendants of immigrants from Turkey in the Netherlands are aware of such attacks, very few respondents from the Russian minorities in Estonia, Latvia and Lithuania are aware of such incidents (2 % or fewer – based on small numbers).
Both in terms of awareness of hate-motivated harass-ment and physical attacks, the largest within-country differences between two interviewed groups are found in Austria and Belgium. For example, in Austria, 29 % of immigrants and descendants of immigrants from Turkey are aware of a family member or a friend who has been insulted or called names because of their eth-nic or immigrant background, compared with 47 % of persons with Sub-Saharan African backgrounds being aware of such incidents. In Belgium, 23 % of immigrants and descendants of immigrants from Turkey are aware of a family member or a friend experiencing insults or name-calling, compared with 40 % of immigrants and descendants of immigrants from North Africa.
KEY FINDINGS
n Of all respondents, 14 % were stopped by the police at least once in the 12 months preceding the survey, and 6 % say the most recent stop was because of their immigrant or ethnic minority background.
n Of those respondents the police stopped in the 12 months before the survey, 40 % say the last stop was because of their immigrant or ethnic minority background – although results vary among target groups and EU Member States.
n In the five years preceding the survey, 26 % of all respondents were stopped by the police and 8 % say that the most recent stop was because of their immigrant or ethnic minority background.
n Of those the police stopped in the past five years, 33 % believe that the most recent stop was because of their immigrant or ethnic minority background – again with significant variations between target groups and EU Member States.
n Among all groups surveyed – similarly to the findings of EU-MIDIS I – respondents with North African and Sub-Saharan African backgrounds indicate being stopped by the police more frequently than other immigrant groups surveyed. Roma respondents indicate being stopped by the police the third most often.
n On average, young respondents indicate being stopped more often than those who are older; and men are stopped much more often than women. These tendencies are more often observed among the respondents with immigrant backgrounds.
n Among Roma respondents, the rate of police stops is quite evenly distributed across different age groups. Roma women and men believe to the same extent that the most recent police stop they experienced was of discriminatory nature.
n Both EU-MIDIS I and EU-MIDIS II results show that respondents of the Russian ethnic minority generally do not believe that they were stopped because of their ethnic origin.
n During the most recent police stop they experienced, a majority of the respondents (58 %) were stopped while in a private car; a quarter (26 %) were stopped on the street while on foot. Roma respondents and respond-ents with Sub-Saharan African background are stopped on the street more often than respondents from other target groups.
2�3� Police stops
69
What do the results show?
Discriminatory misconduct – such as physical assault – and discriminatory ethnic profiling – stops based solely or mainly on individuals’ personal characteristics rather than on their behaviour – by law enforcement are not just unlawful. Such acts can undermine trust in law enforcement and damage community relations, ultimately undermining public cooperation with, and hampering, law enforcement efforts.
2�3�1� Encounters with law enforcement
The results concerning police stops refer to contacts between law enforcement and the survey respondents. Respondents were also asked if they thought that the most recent police stop had happened because of their immigrant or ethnic minority background and about the way they were treated by the police.
The results show that 26 % of all EU-MIDIS II respondents were stopped by the police in the five years before the survey, and that 8 % believe that the most recent stop occurred because of their ethnic minority or immigrant background. Of those stopped in the five years before the survey, 33 % say that the most recent stop occurred because of their ethnic or immigrant background.
When asked about the 12 months preceding the sur-vey, 14 % of all respondents said they were stopped by the police, and 6 % indicated that the most recent stop was because of their immigrant or ethnic minority background. Of those stopped during this timeframe, 40 % believe that the most recent stop was because of their ethnic or immigrant background.
The results show significant variations between target groups and Member States. On average, encounters with law enforcement are more common for immigrants and descendants of immigrants from different countries than for ethnic and national minorities. For example, one third of respondents with Turkish (31 %) and North African (30 %) backgrounds, and one fourth of recent immigrants (27 %) as well as immigrants and descend-ants of immigrants from Sub-Saharan Africa (24 %) said they have been stopped by the police in the five years before the survey. By contrast, 19 % of Roma respond-ents and 14 % of respondents of the Russian ethnic minority say this (Figure 26). Meanwhile, on average, immigrants and descendants of immigrants from Asia and South Asia were stopped the least often (12 %).
When comparing the five-year police stop rates with the 12-month rates, the aggregated results on the target group-level show that the same groups are stopped more often, with immigrants and descendants of immi-grants from North Africa and Turkey being stopped the most often.
A look at different target groups across different EU Member States reveals that certain groups are stopped particularly often in certain countries. For example, more than half of respondents with Sub-Saharan Afri-can background in Austria (66 %) and with South Asian background in Greece (60 %) were stopped by the police in the five years before the survey. Nearly half of respondents with Asian background in Cyprus (42 %), Turkish background in Austria (41 %), and Sub-Saharan African background in Finland (38 %) were stopped by the police in the five years before the survey. Nearly half of Roma respondents had similar experiences in Greece (48 %), Spain and Croatia (46 % and 45 %, respectively). One third of respondents with Sub-Saharan African background in Germany (34 %), Luxembourg (30 %), and France (29 %), and respondents with North African background in the Netherlands (33 %), Italy (32 %), and France (31 %) were stopped in the five years before the survey. Finally, one third of Roma respondents were stopped by the police during this period in Portugal (34 %) and Hungary (33 %) (Figure 26).
Among all respondents with Sub-Saharan African back-ground, on average, 10 % were both stopped in the five years before the survey and believed they were stopped because of their immigrant or ethnic minor-ity background. This was also the case for 11 % of all respondents with North African background and 8 % of all Roma respondents. For respondents with Turkish background, the rate is 5 %. Meanwhile, none of the respondents from the Russian minority perceived as discriminatory the most recently experienced police stop in the five years before the survey.
When looking at the 12-month timeframe, respondents from the same target groups – namely immigrants and descendants of immigrants from North Africa, Sub-Saharan Africa and Turkey, as well as Roma respond-ents – more often than other groups say that the most recent police stop was of discriminatory nature (9 %, 5 %, 4 %, and 6 %, respectively).
EU-MIDIS II findings show that, within some Member States in which more than one target group was sur-veyed, the prevalence of perceived ethnic profiling in the five years before the survey varies among the dif-ferent groups. For example, in Austria, the prevalence of perceived ethnic profiling for respondents with Sub-Saharan African background is 37 %, while for respond-ents with Turkish background it is 6 %. A similar pattern can be noted for respondents with North African and Turkish backgrounds in Belgium (13 % vs 5 %) and in the Netherlands (20 % vs 13 %). By contrast, respond-ents with Sub-Saharan African and North African back-grounds in France perceive nearly the same levels of ethnic profiling (12 % vs 10 %). In Italy, close levels of perceived profiling can be noted between respondents
Second European Union Minorities and Discrimination Survey – Main results
70
Figure 26: Prevalence of stops by police in 5 years before the survey, by EU Member State and target group (%) a,b,c,d,e,f
Stopped, but no ethnic profiling Not stoppedStopped, with perceived ethnic profiling
4020 806010 5030 90700
EE
NL
HU
Group Average
AT
UK
MT
FR
AT
LT
Group Average
PT
BG
BE
Group Average
PT
IE
DE
LV
RO
BE
CZ
DE
SE
IT
DK
Group Average
CY
SK
ES
EL
DK
PL
UK
LU
FI
Total
IT
IT
HR
SE
Group Average
EL
Group Average
FR
ES
NL
SI
Group Average
TUR
NOAFR
(S)ASIA
RIMGR
ROMA
RUSMIN
EU-28
SSAFR
100
14 20 66
2810 6112 17 71
(5) 13 8217 11 72
10 14 76
12 18 69(5) (3) 92
7 12 82
7 11 837 7 86
10 11 79
37 29 34
6 35 59
5 27 68(5) 23 73
13 16 71(5) 10 855 26 68
5 16 79
32
18 24 5753 7 41
30 18 5321 25 5520 25 55
28 (6) 66
126 82
11 89
118 80
15 8516 84
14 86
9 24 67
12 9 796 92
2 2 96
25 73
10 12 78
13 8666
4 5 916 6 88
13 12 7511 12 77
10 21 6923 9 68
20 13 6711 19 70
188 74
(1)
(2)(2)
(1)
Notes: a Out of all respondents (n=25,515); weighted results. b The total percentage of respondents who were stopped by the police in the five years before the survey is calculated
by adding together two figures: the percentage figure of those who were stopped by the police in the five years before the survey and perceived that this was because of their immigrant or ethnic minority background, and the percentage figure of those who were stopped by the police in the five years before the survey, but did not consider that this was because of their immigrant or ethnic minority background.
c Results based on a small number of responses are statistically less reliable. Thus, results based on 20 to 49 unweighted observations in a group total or based on cells with fewer than 20 unweighted observations are noted in parentheses. Results based on fewer than 20 unweighted observations in a group total are not published.
d Question: “In the past five years in [COUNTRY] (or since you have been in [COUNTRY]), have you ever been stopped, searched or questioned by the police?”
e Acronyms for target groups refer to immigrants from [country/region] and their descendants: TUR = Turkey, SSAFR = Sub-Saharan Africa, NOAFR = North Africa, SASIA = South Asia, ASIA = Asia, RIMGR = recent immigrants from non-EU countries, RUSMIN = Russian minority, ROMA = Roma minority.
f Some bars do not add up to 100 %; this is due to rounding of numbers.Source: FRA, EU-MIDIS II 2016
71
What do the results show?
with Sub-Saharan and North African backgrounds (17 % vs 23 %), while this rate is lower for respondents with Asian background (10 %) in the country.
Figure 26 shows differences in experiences of perceived ethnic profiling among target groups and Member States. For example, the highest perceived ethnic profil-ing prevalence rates for respondents with Sub-Saharan African background are observed in Austria. Among respondents with North African background, the high-est rates of perceived ethnic profiling are observed in Italy and the Netherlands. Most respondents with South Asian background felt ethnically profiled in Greece. Finally, Roma respondents felt ethnically profiled in the five years before the survey at the highest shares in Greece, Portugal, Spain and Croatia.
Extent of perceived ethnic profiling among those stopped by the police
A relatively high proportion of the respondents who were stopped by the police in the five years before the survey believe that this was because of their immi-grant or ethnic minority background. On average, nearly every second respondent with Asian background (47 %) and Sub-Saharan (41 %) and North African (38 %) backgrounds who were stopped during this time-frame said that this was due to this reason. Similarly, of stopped Roma respondents, nearly every second (42 %) believed this was because of their ethnic back-ground. By contrast, this share is much lower among the stopped respondents with Turkish background (17 %) (Figure 27). The target groups fall into the same order with regard to perceived ethnic profiling in the 12 months preceding the survey: stopped respondents with Asian background indicated the highest rates and those with Turkish background the lowest (the rates range from 61 % to 26 %, respectively).
When looking at specific target groups’ experiences with discriminatory stops in different EU Member States, these vary across the Member States. Fig-ure 27 shows that stopped respondents with Asian background in Greece indicated the highest rates of perceived discriminatory police stops in the five years before the survey, with much lower rates in Italy and the United Kingdom. Among the stopped respondents with Sub-Saharan African background, the highest rates of perceived discriminatory police stops were indicated in Italy, Austria, and Sweden. Among respondents with North African background who were stopped, the high-est rates were mentioned in Italy, the Netherlands, and Belgium. Finally, among Roma respondents who were stopped, the rates of perceived ethnic profiling are higher than the aggregated group average in Portugal, Greece, the Czech Republic and Romania.
Overall, the EU-MIDIS II results indicate that discrimina-tory police practices affect certain respondent groups more than others, which is consistent with findings in EU-MIDIS I. In both surveys, immigrants and descend-ants of immigrants from North Africa and Sub-Saharan Africa, as well as Roma respondents, most frequently said that the police stopped them because of their eth-nic or immigrant background. Both EU-MIDIS I and EU-MIDIS II results show that respondents of the Russian ethnic minority generally do not believe that they were stopped because of their ethnic origin.
2�3�2� Differences in police stops by gender and age
A look at gender differences reveals that men are stopped by the police more often than women across all surveyed target groups. The greatest differences between men and women who were stopped are observed among immigrants and descendants of immigrants from North Africa (48 % vs 13 %) and Sub-Saharan Africa (39 % vs 11 %), followed by immigrants and descendants of immigrants from Turkey (48 % vs 14 %) and Roma respondents (29 % vs 12 %).
Gender differences are also observed between several target groups in terms of perceptions of the discrimi-natory nature of police stops. Among stopped Roma respondents, men and women on average perceived profiling on equal levels, with 42 % of both genders believing that the most recent police stop in the five years before the survey occurred because of their eth-nic background. Among immigrants and descendants of immigrants from Sub-Saharan Africa, of those stopped, similar levels perceived profiling (men: 44 %, women: 34 %). Among stopped respondents with North Afri-can background, on average, more men than women believed this was because of their immigrant or ethnic background (44 % and 15 %, respectively) (Figure 28).
The survey data also reveal differences across the Member States. Among respondents with North Afri-can background stopped by the police in the five years before the survey, the majority of male respondents in Italy, the Netherlands, and Belgium (74 %, 66 %, and 60 %, respectively) believed that the last stop was of a discriminatory nature. Among stopped respondents with Sub-Saharan African background, half of male respondents in Austria, Italy, Sweden, and Denmark (59 %, 56 %, 54 %, and 49 %, respectively) believed that the last stop was because of their immigrant or ethnic origin. Among stopped Roma respondents, more women than men believed they were stopped because of their ethnic background in Portugal and Greece (96 % vs 77 % and 68 % vs 59 %, respectively).
Second European Union Minorities and Discrimination Survey – Main results
72
Figure 27: Most recent police stop being perceived as ethnic profiling among those who were stopped in 5 years before the survey, by EU Member State and target group (%) a,b,c,d
4020 806010 5030 90700
EE
FR
HR
Group Average
NL
CY
UK
DK
MT
LT
Group Average
SK
PT
SE
Group Average
PT
DE
IT
LV
HU
IT
EL
BE
IE
FR
AT
Group Average
EL
BG
NL
CZ
DK
PL
FI
LU
SE
Total
UK
ES
ES
AT
Group Average
IT
Group Average
BE
RO
DE
SI
Group Average
TUR
NOAFR
(S)ASIA
RIMGR
ROMA
RUSMIN
EU-28
SSAFR
100
33
(66)60
5651
47
4240
3835
2827
41
43(31)
23(17)
1614
17
71
8946
(9)(5)
8463
5752
4645
3527
1942
0000
33
(6)
45
4743
6152
47
38
42
Notes: a Out of respondents who were stopped by the police in the five years before the survey (n=6,787); weighted results. b Results based on a small number of responses are statistically less reliable. Thus, results based on 20 to 49 unweighted
observations in a group total or based on cells with fewer than 20 unweighted observations are noted in parentheses. Results based on fewer than 20 unweighted observations in a group total are not published.
c Questions: “In the past five years in [COUNTRY] (or since you have been in [COUNTRY]), have you ever been stopped, searched or questioned by the police?”; “Do you think that THE LAST TIME you were stopped was because of your ethnic or immigrant background?”
d Acronyms for target groups refer to immigrants from [country/region] and their descendants: TUR = Turkey, SSAFR = Sub-Saharan Africa, NOAFR = North Africa, SASIA = South Asia, ASIA = Asia, RIMGR = recent immigrants from non-EU countries, RUSMIN = Russian minority, ROMA = Roma minority.
Source: FRA, EU-MIDIS II 2016
73
What do the results show?
Young respondents are more frequently stopped. Every third young respondent with North African, Turkish and Sub-Saharan African backgrounds was stopped by the police in the five years before the survey: 37 %, 30 %, and 37 % of those aged 16 to 24, respectively; and 39 %, 40 %, and 30 % of those aged 24 to 34, respectively. The police stop rate among older respondents is closer to the target group average, with stops less prevalent for the oldest age groups (e.g. 55+).
Among Roma respondents, the rate of police stops is quite similar across age groups. Nearly every fifth Roma was stopped by the police in the five years before the survey: 24 % of those aged 16 to 24; 22 % of those aged 24 to 34; 21 % of those aged 35 to 44; and 20 % of those aged 45 to 54. Among Roma aged 55 and above, the rate is 11 %. However, Roma in this group are distinct in that they believe at a higher rate that the last stop by the police in the five years before the survey was based on ethnic profiling (51 % of those aged 55 to 64, compared to 41 % to 44 % across other age groups).
Figure 28: Most recent police stop being perceived as ethnic profiling among those who were stopped in 5 years before the survey, by gender and target group (%) a,b,c,d
0
40
20
80
60
10
50
30
100
90
70
SSAFR TUR NOAFR (S)ASIA ROMA RIMGR Total
19
3733
(4) (6) (6)
42 42 42
35
48 47
15
4438
(5)
2117
34
4441
Women Men Group averageNotes: a Out of respondents who were stopped by the police in the five years before the survey (n=6,787); weighted results. b Results based on a small number of responses are statistically less reliable. Thus, results based on 20 to
49 unweighted observations in a group total or based on cells with fewer than 20 unweighted observations are noted in parentheses. Results based on fewer than 20 unweighted observations in a group total are not published.
c Questions: “In the past five years in [COUNTRY] (or since you have been in [COUNTRY]), have you ever been stopped, searched or questioned by the police?”; “Do you think that THE LAST TIME you were stopped was because of your ethnic or immigrant background?”
d Acronyms for target groups refer to immigrants from [country/region] and their descendants: TUR = Turkey, SSAFR = Sub-Saharan Africa, NOAFR = North Africa, SASIA = South Asia, ASIA = Asia, RIMGR = recent immigrants from non-EU countries, ROMA = Roma minority.
Source: FRA, EU-MIDIS II 2016
Second European Union Minorities and Discrimination Survey – Main results
74
2�3�3� Circumstances and nature of most recent police stop
The survey interviews were conducted during a time period that included major terrorist attacks in France and Belgium,38 which prompted an increase in police surveillance and identity checks. Similarly, migration movements through Greece and Italy also spurred increased police and border checks.
The survey asked respondents about the circumstances of the most recently encountered police stop, asking whether they were stopped while in a private car, on public transport or on the street. Of those respond-ents in the 28 EU Member States who were stopped, a majority (58 %) were stopped while in a private car; the highest rates are observed among the Russian minor-ity (75 %), recent immigrants (73 %), and respondents with Turkish (72 %) and North African backgrounds (59 %). On average, this share is lower among stopped respondents with Asian (49 %) and Sub-Saharan African (44 %) backgrounds, as well as among stopped Roma respondents (34 %).
On average, of those stopped, nearly one in four (26 %) respondents were most recently stopped by the police on the street while on foot, and 5 % said they were stopped while travelling on public transport. The results indicate that certain groups are stopped on the street more often than others, with significant variations between Member States. On average, among those stopped, nearly every second Roma respondent (44 %), every third respondent with Sub-Saharan African back-ground (34 %), and every fourth respondent with North African background (27 %) was stopped on the street. Most Roma who were stopped in the Czech Republic, Spain and Slovakia (64 %, 51 %, and 46 %, respec-tively) were most recently stopped by the police on the street. Half of stopped respondents with Sub-Saharan African background in Austria and Luxembourg39 (57 % and 51 %, respectively) were stopped on the street. Among stopped respondents with North African back-ground, nearly half were stopped on the street in the Netherlands, Italy and Spain (44 %, 41 %, and 40 %, respectively). The majority of stopped respondents with South Asian background in Greece (79 %) were stopped by the police on the street, which could be related to intensive immigration checks or to the fact that most immigrants in Greece are younger and represent more recent arrivals and, therefore, are not very likely to possess/use a car.
38 On 14 July 2016, a lorry was driven into crowds in Nice, France, resulting in 86 deaths and 434 injuries. On 22 March 2016, three suicide bombings in Brussels, Belgium – at the airport and at a metro station – resulted in 32 deaths and more than 300 injuries.
39 Results for Luxembourg should be interpreted with caution due to the sampling design applied.
The results show that, when stopped by the police, most respondents in the EU-28 were asked for their identity papers (65 %), driving licence or vehicle docu-ments (51 %), or other questions (51 %). A quarter of those stopped (24 %) say the police searched them or their car. 13 % of all stopped respondents were fined during the most recent police stop, 12 % were given an alcohol or drug test, 11 % received some form of advice or warning from the police, and 5 % say they were arrested or taken to a police station.
With regard to identity checks, almost all stopped respondents with South Asian background in Greece (96 %), with Sub-Saharan African background in Austria (99 %), Italy (96 %) and Luxembourg40 (90 %), as well as with North African background in Italy (92 %) were asked for ID cards, passports or residence permits dur-ing the latest police stop. This can be explained by the migration flows during that period in both countries.
More than eight in 10 immigrants and descendants of immigrants from Turkey and North Africa who live in Belgium (86 % and 84 %, respectively) and recent immigrants in Poland (89 %) who were stopped were asked to provide their identity papers during the latest police stop. Similar shares of stopped Roma respond-ents in Spain, the Czech Republic and Croatia (83-88 %) were asked to provide their identity papers during the latest police stop.
The highest so-called ‘hit rates’ resulting from police stops – i.e., the proportion of stops and searches that resulted in law enforcement sanctions, such as a fine, apprehension or traffic ticket – were among stopped respondents of Turkish origin in Austria and stopped Roma respondents in Portugal: half of each reported being fined (50 % and 51 %, respectively). In case of stopped respondents of Turkish origin in Austria, the majority were also asked for their driving licence or vehi-cle documents (83 %) or their identity papers (54 %). Among stopped Roma respondents in Portugal, close to half were also asked for identity papers (59 %) or their driving licence or vehicle documents (48 %). How-ever, though many stopped respondents with Turkish background were sanctioned as a result of their most recent stops in Austria, only 14 % of those stopped considered the stop to have been discriminatory. In Por-tugal, the majority of stopped Roma (84 %) considered the stop discriminatory.
2�3�4� Treatment by police during stops
A majority (59 %) of all respondents who were stopped by the police in the five years before the survey note
40 Results for Luxembourg should be interpreted with caution due to the sampling design applied.
75
What do the results show?
that they were treated respectfully (25 % ’very respect-ful’, 34 % ‘fairly respectful’). One in four (24 %) said the way police treated them was ‘neither respectful, nor disrespectful’. Meanwhile, 17 % said that the police treated them disrespectfully (8 % ‘fairly disrespectfully’ and 9 % ‘very disrespectfully’).
Roma respondents and respondents with North African background who were stopped more often indicated experiencing disrespectful behaviour by police during the most recent stop (25 % and 21 %, respectively) than other target groups. The majority of stopped Roma respondents in keep together on one line referred to disrespectful treatment by the police. Nearly half (43 %) of stopped respondents with North African background in the Netherlands believe that they were treated dis-respectfully during the most recent police stop. Every third stopped respondent with Sub-Saharan African background in Malta and Denmark, and with Turkish background in the Netherlands, noted disrespectful treatment by the police.
Respondents in Italy who were stopped distinctly assess the police’s behaviour during the most recent stop: 56 % of stopped respondents with North Afri-can background, 37 % with Sub-Saharan African back-ground, and 34 % with Asian background selected the response ‘neither respectful, nor disrespectful’. Some 42 % of respondents with Sub-Saharan African back-ground in Austria held the same opinion.
Among those stopped, the majority of recent immi-grants in Slovenia (94 %) and Poland (85 %), as well as respondents with Russian minority background in Lithuania (91 %) and Latvia (74 %), indicated that they
were treated respectfully during the most recent police stop. Similar shares of stopped Roma respondents in Hungary (73 %), and stopped respondents with Turkish background in Sweden (73 %) and Belgium (71 %), were of the same opinion.
Of all stopped respondents who indicated that they were treated disrespectfully during the latest police stop, 7 % reported or made a complaint about the police’s inappropriate behaviour. The majority of cases were not reported.
As results presented in Section 2.4.5 show, on average, respondents across the target groups – except for Roma – tend to trust the police and the local (municipal) authori-ties where they live more than other institutions asked about in the survey: the overall average is 6.3 on a scale where zero means ‘no trust at all’ and 10 means ‘com-plete trust’. On average, the lowest levels of trust in the police are observed among Roma respondents in Spain, Slovakia, the Czech Republic, Portugal, and Bulgaria – ranging from 3.7 to 4.6. (See Figure 42.) Although other target groups are quite close to the overall average, respondents with Sub-Saharan African background in Austria are least trustful of the police (3.6). Respondents with North African background in the Netherlands, Italy and Belgium, and with Sub-Saharan African background in France, tend to have relatively little trust in the police (averages range from 5.0 to 5.6). The highest levels of trust in the police are observed among respondents with Sub-Saharan African background in Finland, Malta and Germany; with Turkish background in Germany, Aus-tria and Denmark; with South Asian background in the United Kingdom; and recent immigrants in Slovenia (with averages ranging from 6.9 to 8.2).
Second European Union Minorities and Discrimination Survey – Main results
76
2�4� Living together: residence and citizenship, socio-economic situation, trust and tolerance
KEY FINDINGS – RESIDENCE AND CITIZENSHIP
n First-generation immigrants hold a variety of residence statuses and their access to rights varies accordingly. Most immigrants benefit from a secure residence status (74 %), which means they either hold national citizen-ship or have residence permits that are valid for at least five years. Immigrants who have been in a country for a longer time naturally more often have a secure residence status. However, immigrants from Sub-Saharan Africa in Austria (47 %) and Portugal (45 %) often do not hold a secure residence status even after residing in the respective country for 10 years or longer.
n The majority of respondents hold national citizenship (women: 60 %, men: 54 %), with strong differences across countries and target groups. Among first-generation immigrants, slightly more than 42 % hold national citizenship; women hold citizenship more often than men in this group (women: 45 %, men: 40 %), with stronger gender differences observable among Sub-Saharan African immigrants in Denmark, Finland, Ireland and Sweden. Among second-generation respondents, 87 % hold national citizenship.
n Among first-generation immigrants, the most important reasons for not applying for citizenship are that they do not fulfil the requirements (30 % of those who have not applied) and that they do not want to give up their current citizenship (24 %).
n Close to all descendants of immigrants – the second generation – have a secure residence status. However, some cannot or choose not to acquire citizenship of their country of residence; the latter is particularly the case among second-generation respondents from Turkey (with 67 % holding national citizenship).
n Overall, 7 % of all respondents who are immigrants and descendants of immigrants have close family mem-bers who are living outside their country of residence, but this rate varies considerably across groups and countries. These family members may be able to join them via family reunification channels. Immigrants and descendants of immigrants with recent immigration histories tend to have larger shares of close family mem-bers outside the country – such as immigrants from South Asia in Greece (44 %) and immigrants from Asia in Cyprus (29 %). Among recent immigrants in Slovenia, one in five have family outside the country (21 %). The same rate can be found among immigrants from Sub-Saharan Africa in several countries, including Austria, Italy, Luxembourg and Malta (ranging between 19-20 %).
n One in five respondents with close family members living outside their country of residence have previously applied for family reunification; one third had their applications rejected. One in three respondents with family outside the country who have not applied for family reunification have not done so because they cannot fulfil the requirements, could not afford to sustain their family members if they were to join them, do not want to go through the procedure or find the fees too expensive.
KEY FINDINGS – SOCIO-ECONOMIC SITUATION
n Among EU-MIDIS II respondents aged 16 to 64, a lower proportion (61 %) had completed at least upper sec-ondary education in 2016 than among the general population aged 15 to 64 (74 %). While there are differ-ences between countries and target groups, in general, second-generation respondents have obtained higher education levels than first-generation respondents – except in Germany, where there is almost no difference between first- and second-generation respondents of Turkish descent.
77
What do the results show?
n Among respondents aged 16 to 64 who had obtained at most a lower secondary education (ISCED 0-2), 38 % were no longer pursuing any education or training at the time of EU-MIDIS II. This share decreases for re-spondents of younger age and for second-generation respondents. Among the target groups surveyed in EU-MIDIS II, Roma respondents report having completed the lowest levels of education, and indicate at the highest rates that they were not pursuing further education.
n On average, 71 % of all respondents say that they are proficient in speaking, reading and writing the national language of their country of residence. For the second generation, this increases to almost 100 %, regardless of their parents’ country of origin or their country of residence. In general, men indicate that they have a higher language proficiency than women do.
n While the overall ability to speak, read and write in a country’s national language(s) provides a strong basis for integrating into its society (including the labour market), the ability to read in that language gives people the opportunity to learn about their rights. On average, 9 % of respondents are insufficiently proficient in the national language in terms of reading. As many as 4 % of respondents cannot read in the language of the coun-try in which they live (where they were interviewed) at all. This means that they have no opportunity to learn about their rights if such information is available only in the national language. For most groups surveyed in EU-MIDIS II, language proficiency increases with educational levels; however, for the Russian minority, higher educational levels do not correspond to higher proficiency in the respective national language.
n On average, 61 % of working-age people living in respondents' households indicate that they were employed or had done some work in the four weeks preceding the survey. This is below the average EU-28 employment rate of 71 % in 2016. A large gender gap can be observed, with women less often engaged in paid work, across target groups and countries – except for persons with Sub-Saharan African background in Austria and Portugal, persons from Turkey in Sweden, persons from Asia in Cyprus, and members of the Russian minority in Estonia. The gender gap among the various target groups is, on average, almost twice as high as among the general population. Across all target groups, no differences could be observed between the paid work rates of first-generation and second-generation respondents.
n Attaining a higher educational level positively influences respondents’ position on the labour market. On av-erage, only 52 % of respondents who completed at most lower secondary education report being in paid work, while 67 % of those with at least upper secondary education do so. Completing more than just lower secondary education influences the labour market participation of Roma respondents particularly positively in the Czech Republic, followed by recent immigrants; respondents with North African background in the Nether-lands; and with Turkish background in Sweden, Denmark and the Netherlands.
n Respondents’ ability to participate in the labour market is also related to their proficiency in the national lan-guage. On average, only 43 % of those with insufficient language proficiency in all three dimensions (speak-ing, reading and writing) report being in paid work in the four weeks before the survey. This share increases to 65 % for respondents who are sufficiently proficient in all three dimensions. Language proficiency has a particularly positive influence on the rate of being engaged in paid work for respondents with Turkish back-ground – across all countries in which they were interviewed except for Austria. For this group, high language proficiency increases their paid work rate by more than 100 % (on average, from 35 % to 78 %).
KEY FINDINGS – TRUST AND TOLERANCE
n Among different institutions, respondents have the most trust in local authorities. This means that local au-thorities are well placed to provide integration measures for immigrants, descendants of immigrants, and ethnic minorities. There is also a very high level of trust in the police and the legal system, but less trust in politicians and political parties – in line with the general population. Levels of trust are often very similar to, or even higher than, the general population’s trust levels. Among immigrants, trust is lower among the second generation in almost all countries and target groups, but especially in France and the United Kingdom. In some countries, Roma indicate lower levels of trust in the police than the general population.
Second European Union Minorities and Discrimination Survey – Main results
78
Participation in the democratic process as active citizens supports the integration of immigrants, descendants of immigrants, and ethnic minorities, and enhances their sense of belonging. This section analyses questions relating to integration and social inclusion – looking at issues such as residence and citizenship, socio-eco-nomic situation, as well as trust and tolerance. These issues affect the diverse groups covered in the sur-vey in different ways. For example, some are relevant for first-generation immigrants, but not for persons born in their country of residence. Furthermore, dif-ferent factors play a role for immigrants and ethnic minorities when it comes to outcomes related to the socio-economic situation and issues related to trust and tolerance. Most importantly, the challenges Roma communities nowadays face with respect to their social inclusion in European societies often differ from the challenges immigrants, descendants of immigrants or Russian minorities face. FRA therefore examined the
situation of Roma in a November 2016 report focusing on EU-MIDIS II data relating to their specific experiences.
Given that immigrants’ migration history and legal situ-ation have particular implications for their access to a number of rights, this section first provides an overview of first-generation immigrants’ residence and citizen-ship status. The second part focuses on the socio-eco-nomic situation – education, language proficiency and labour market participation – of all target groups, includ-ing first- and second-generation respondents, Roma and Russian minorities. The third section discusses the level of trust in public institutions and the level of attachment to, and identification with, the survey country, as well as opinions related to tolerance and participation. Finally, this section also looks at possible effects of discrimination and bias-motivated victimisa-tion on respondents’ sense of national belonging and their trust in public institutions.
n As among the general population, trust is related to interest in politics – with respondents expressing high-er trust in institutions also showing more interest in politics. Respondents with national citizenship indicate higher levels of interest in politics.
n The majority of respondents feel strongly attached to their country of residence and many identify with their country of residence. Both first- and second-generation respondents also indicate that they strongly identify with their country of origin. Every second respondent in these target groups identifies with both the country of residence and the country of origin. The second generation indicates stronger feelings of identification with the country of residence. Citizenship is one of the main factors related to an increased sense of belonging to and identifying with the country of residence in the EU. Experiences with discrimination lead to considerably lower feelings of belonging.
n There is a high level of acceptance of other ethnic groups – 95 % feel comfortable with or neutral about having neighbours of a different ethnic origin. This is not the case with respect to LGBT people: only 73 % feel com-fortable with or neutral about having lesbian, gay or bisexual people as neighbours, and 66 % feel comfortable with or neutral about transgender and transsexual neighbours. Many respondents have friends with other or no ethnic minority backgrounds (77 % and 82 %, respectively). Those with such friends indicate higher levels of acceptance of other groups. The positive association between having friends of different ethnic origins and levels of acceptance of other groups is not only observed for EU-MIDIS II respondents, but also for the general population. Acceptance levels of gender equality are high in most groups – with some exceptions, such as im-migrants from South Asia in Greece.
n Respondents who experienced any kind of discrimination, harassment or violence are less likely to identify with their country of residence. Such experiences have a strong effect on respondents’ feelings of attach-ment and identification. While most first-generation immigrants identify equally strongly with their country of residence and their country of birth, discrimination experiences make them identify less with their country of residence. Holding national citizenship also increases respondents’ sense of identifying with the country of residence compared to the country of origin. Additionally, a longer term of residence also contributes to stronger feelings of identification with the country of residence.
n Experiences with discrimination, harassment or violence have a strong effect on the level of trust in the police and the legal system. A lower average level of trust in the police and the legal system for those who experi-enced discrimination, harassment or violence compared to those who did not experience such an event can be observed across almost all target groups, countries, genders and age groups.
79
What do the results show?
2�4�1� Residence status and citizenship
Residence and citizenship status are of major impor-tance for policy developments because they affect the level of access to rights. Individuals without residence status and those holding residence permits of limited validity do not enjoy access to as many rights as those with unlimited residence or especially those holding national citizenship. While the EU has no direct compe-tence on regulations for granting national citizenship, there is EU legislation governing residence status and citizens’ rights. For example, the long-term residence directive41 provides some additional rights to third-country nationals living in the EU. Only those who obtain national citizenship in any of the EU Member States can access all rights related to EU citizenship. Finally, fam-ily reunification is covered by the family reunification directive,42 which governs access rights for family mem-bers of third-country nationals and EU citizens in the EU.
EU-MIDIS II covers the main immigrant groups in the EU. These groups are diverse – with different migra-tion histories in terms of the time of and reasons for immigration, and the political circumstances in both the countries of origin and in the destination countries. This section presents findings related to respondents’ resi-dence and citizenship status, which are central for the enjoyment of rights by immigrants and, in some Mem-ber States, also by their offspring. Since not all countries automatically provide citizenship to those born in the country, this aspect is looked at particularly at the level of second-generation respondents. Some of the issues – such as the main reason for migrating and the length of residence – only affect first-generation immigrants, so these particular results mainly refer to the situa-tion of first-generation immigrants (though they also vary substantially with regard to countries of origin). Therefore, the following section mainly summarises results for first-generation immigrants, and looks at their various legal residence statuses, possible changes in status – including naturalisation – since arrival, as well as family re-unification.
Migration history and residence status of immigrants
The immigrants included in the EU-MIDIS II sample come from 121 different countries43 of origin, with most respondents coming from Morocco and Turkey. Other
41 Council Directive 2003/109/EC of 25 November 2003 concerning the status of third-country nationals who are long-term residents, 2003.
42 Council Directive 2003/86/EC of 22 September 2003 on the right to family reunification, 2003.
43 ‘Country of origin’ does not always refer to actual countries, but also to regions that are not fully acknowledged as individual countries and can be seen as parts of other countries. The separately published Technical Report provides a full list of countries and regions of origin included.
important countries of origin include Somalia, Nigeria, Pakistan and Bangladesh. A more detailed overview of countries of origin by EU Member State can be found in Table 8 in Annex II, and by target group in Figure 29. Countries of origin within the different target groups vary strongly between EU Member States, reflect-ing country-specific migration patterns. For example, looking at respondents with Sub-Saharan African back-ground, 60 % of this target group in Austria come from Nigeria, while in Denmark 91 % come from Somalia. In Germany, the countries of origin of immigrants from Sub-Saharan Africa are much more dispersed, with the most important being Eritrea (19 %), Ghana (18 %) and Togo (11 %).
The length of residence (Table 7 in Annex II) and the period of immigration also vary across target groups, as respondents immigrated during different and changing historical, political and legal circumstances and because most EU Member States changed their approaches to regulating immigration in the past decades. Several waves of significant immigration flows to the EU can be identified. They started in the 1960s (e.g. immi-grants from North Africa to France) and 1970s and 1980s (e.g. labour emigration from Turkey to Austria, France, Germany and the Netherlands, among others, under the so-called “guest-worker” system). They con-tinued throughout the 1990s (e.g. Sub-Saharan Afri-can migration to Denmark, Finland and Sweden), and include the more recent waves of immigration in the 2000s (e.g. immigrants from Sub-Saharan Africa to Aus-tria, Germany, Ireland and Italy). Of all first-generation immigrants in the EU-MIDIS II sample, 39 % arrived before the 1990s, 20 % during the 1990s, and 42 % arrived after 2000.
Reasons for immigration: The survey asked first-gen-eration immigrants – respondents born outside the EU – about their reasons for migrating. Over one third of first-generation respondents (35 %), and most often respondents from South Asia (42 %), indicated that they migrated to the EU for family reasons. Employment was given as the second most important reason – mentioned by 28 % of first-generation immigrants, and by 60 % of recent immigrants. Almost one in four immigrants (23 %) indicated that they came to the EU as a child with their parents, especially immigrants from Turkey (37 %). Still looking at all first-generation immigrants together, almost one in 10 came to study or to seek asylum and protection (9 % and 8 %, respectively), particularly immigrants from Sub-Saharan Africa (16 % and 17 %, respectively).44
The reasons for migration are an important indicator of legal channels available for immigrants to come
44 The percentages add up to over 100 % because respondents were allowed to select more than one reason for migration.
Second European Union Minorities and Discrimination Survey – Main results
80
to the EU, and also reflect the gendered nature of specific migration patterns. Among first-generation immigrants, every second woman and one in five men came for family reasons (49 % vs 22 %). Meanwhile, men more often migrated to the EU for employment reasons than women (41 % vs 14 %). Furthermore, more men than women indicated that they migrated for the purpose of seeking protection (10 % vs 5 %).
Status upon arrival: The legal situation related to immigration and residence in the EU varies consid-erably across Member States. Migration laws differ across countries and, in some Member States, also changed several times during past decades. 32 % of first-generation immigrants had a residence or work permit upon arrival. Almost one in five (19 %) came as children with their parents. Another 15 % did not need a residence permit, either because they were nationals of the country of immigration or were EU nationals.
5 % of respondents came as asylum seekers. About one in 10 respondents migrated to the EU with a tour-ist visa (9 %), and another one in 10 arrived without papers (8 %). There are substantial differences across target groups and countries, particularly with respect to asylum seekers, tourists and irregular immigrants.
Among respondents from Sub-Saharan Africa, the per-centage of asylum seekers is much higher – for example, in Austria, Ireland, Denmark and Malta (67 %, 51 %, 46 %, and 46 %, respectively). Among respondents from Turkey in Sweden, one in five were asylum seekers upon arrival (19 %); the percentage of asylum seekers among this target group is much smaller in the other EU Member States covered in this report. One in five respondents from Turkey in Austria indicated that they arrived as tourists (23 %). This is related to the fact that the requirements for immigration were different before the 1990s, when it was easier to immigrate as
Figure 29: Countries and regions of birth of first-generation immigrants, by target group and sample size a,b,c
Number of respondents 200 400 600 Target group: NOAFR RIMGR SASIA SSAFR TUR
Notes: a EU Member States are marked in grey. b Based on respondents (n=12,766). c For the purpose of the survey, target groups from Sub-Saharan Africa in France and the United Kingdom also
included selected countries and regions in the Caribbean. The detailed list of countries included in the sample is available in the separately published Technical Report.
Source: FRA, EU-MIDIS II 2016
81
What do the results show?
tourists and obtain legal residence afterwards. Holding no residence permit upon arrival is mainly observed among respondents from Sub-Saharan Africa in Malta and Italy (47 % and 42 %, respectively). However, by far the highest percentage of immigrants without residence permit upon arrival is observed among respondents from South Asia in Greece (80 %).
Current residence status and change of status upon arrival
Change of status upon arrival: Many immigrants have changed their residence status since their arrival to the EU. Figure 30 shows both the status upon arrival and the current status; the lines signal the specific
Figure 30: Status upon arrival (left column %), current residence status (right column %) and changes of status (thickness of lines) of immigrants a,b,c
5
3
18
10
18
3
42
6
8
9
6
5
19
32
15
A residence/work permit
Parents' papers
No need for papers
Status upon arrival Current status
Other
No permit
Tourist visa
Student visa
Asylum seeker
Other
No residence permit
Limited permit < 5 years
Limited permit > 5 years
Unlimited residence permit
EU citizen
National citizen
Notes : a Out of all respondents who are first-generation immigrants (upon arrival: n=12,390; current status: n=12,766); weighted results.
b Based on questions: “Did you have any type of permit, documents or status when you FIRST arrived in [COUNTRY]? If yes, which one from the following list did you have?”; “Do you currently have a permit to stay in [COUNTRY]?”; “Altogether, for how long is your current permit valid?”
c Thickness of lines indicates the share of respondents reporting the different residence status upon arrival and current status. It shows the flow from arrival status to current status.
Source: FRA, EU-MIDIS II 2016
Second European Union Minorities and Discrimination Survey – Main results
82
change in status over time.45 More than half of the respondents who did not need a residence permit upon arrival or who came on their parents’ papers are cur-rently national citizens (64 %). One third of those who came with a residence or work permit also became national citizens (33 %), and another third of respond-ents (29 %) currently holds a residence permit with unlimited validity. Of those who came with residence or work permits, 32 % hold a residence permit with limited validity, mainly with validity of less than five years. Almost half of the respondents who came as asylum seekers have in the meantime become national citizens (47 %); however, 7 % indicate that they cur-rently do not hold a valid residence permit. The major-ity of those who arrived as tourists are currently either national citizens (35 %) or hold a long- or short-term residence permit (54 %). 7 % of those who came as tourists currently do not hold a residence permit. The majority of those who say that they arrived without a residence permit currently hold a legal status by either holding national citizenship (23 %) or a residence permit (61 %); only 7 % still (or again) lack a residence
45 The categories differ because it was not possible to ask very detailed questions about the status upon arrival, given that many respondents did not know or did not remember the details asked about.
permit. These findings indicate that a change in resi-dence status over time is quite common among the different target groups surveyed in EU-MIDIS II.
Current residence status: Nearly half of first-generation respondents currently hold EU citizenship, including the citizenship of their country of residence or of another EU Member State (42 % and 3 %, respectively). 18 % hold a residence permit with unlimited validity. Another 10 % hold a permit with limited validity that is valid for more than 5 years. 18 % hold a residence permit that is valid for less than 5 years. Overall, 3 % of all first-generation respondents do not currently hold a residence permit. The remaining 5 % are currently in the process of renew-ing their residence permit – or indicated that they did not know if they had a residence permit or refused to answer this specific survey question.
There are important differences in the current residence statuses across the target groups as well as across the Member States in which these same target groups were surveyed. Figure 31 shows the current residence
Figure 31: Current legal residence status among first-generation immigrants, by target group (%) a,b
0
40
20
80
60
10
50
30
100
90
70
SSAFR TUR NOAFR (S)ASIA RIMGR Total
54
18
984
52
51
5
53
3
32
43
28
20
43
38
64
13115
62
10
2
29
6
43
953
18
10
18
3
42
National citizen
Limited residence permit > 5 years
Other (e.g. currently renewing, don’t know)
EU citizen, but not national
Limited residence permit < 5 years
Unlimited residence permit
No residence permit
Notes: a Out of all respondents who are first-generation immigrants (current status: n=12,766); weighted results. b Based on questions: “Do you currently have a permit to stay in [COUNTRY]?”; “Altogether, for how long is your
current permit valid?”Source: FRA, EU-MIDIS II 2016
83
What do the results show?
status of first-generation immigrants at the level of the aggregate target groups.
Respondents with national citizenship or with a residence permit valid for more than five years have a more secure status compared to respondents with residence per-mits of limited validity or those without any permit. The threshold of five years for deeming a residence permit ‘secure’ was chosen because five years of legal residence would allow access to a long-term residence status under EU law, if certain nationally stipulated requirements are met. This definition of a “secure residence status” is not clear-cut, but provides an indication of the percentage of respondents with secure residence status.
Across all first-generation respondents surveyed in EU-MIDIS II, three in four have secure residence (74 %). Differences across target groups and Member States are again observed (Figure 32). As expected, the rate is lower among recent immigrants, although a higher share of this group holds a long-term residence permit or national citizenship in Slovenia (69 %) than in Poland (38 %). Among respondents from Sub-Saharan Africa, the vast majority of those living in Ireland (87 %) and in the United Kingdom (86 %) hold the citizenship of their country of residence or hold a long-term resi-dence permit; this percentage is considerably lower
among the same target group in Malta (2 %), Aus-tria (17 %), Portugal (39 %) and Italy (40 %). Among respondents from Turkey, almost 9 in 10 (88 %) have a secure residence, with higher shares in Denmark and the Netherlands (92 %) and comparably lower shares in Austria (77 %) and Belgium (76 %).
The situation of respondents from North Africa also differs across Member States, with more stable resi-dence statuses observed in the Netherlands (91 %) and low shares of secure residence in Spain and Italy (26 % and 38 %, respectively). Meanwhile, the major-ity of immigrants from South Asia hold secure status in the United Kingdom (86 %), but less than a third do so in Greece (29 %).
These differences are strongly influenced by the dura-tion of residence and partly by the migration channels available for immigration to the EU from different coun-tries of origin. To account for this, Figure 32 includes bars indicating the percentage of immigrants with citizen-ship or residence permits valid for five years or more after at least 10 years of residence. Looking at those with 10 years of residence or more increases the per-centage of those with citizenship or long-term residence status considerably – to 82 % of first-generation immi-grants. Still, slightly less than half of Sub-Saharan African
Figure 32: First-generation respondents with secure residence status, by length of residence, target group and country (%) a,b
All immigrants Immigrants with residence of 10 years or longer
0
40
20
80
60
10
50
30
100
90
70
AT DE DK FI FR IESSAFR TUR NOAFR (S)ASIA RIMGR
IT LU MT PT SE UK AT BE DE DK NL SE BE ES FR IT NL CY EL IT UK PL SI
17
47
81
91
81
92
78
94
68
78 87
40
51
(2)
39
75
86
77 76
8992 92
87
71
26
73
38
91
3529
46
86
38
69
87
5760
45
9693
81 8192
96 94 96
80
29
80
50
92
5549 49
97
(70)
99
Notes: a All first-generation immigrants (n=12,766) and immigrants with at least 10 years residence (n=8,064). b Based on questions: “Which citizenship do you have?”; “Do you currently have a permit to stay in [COUNTRY]?”;
“Altogether, for how long is your current permit valid?”Source: FRA, EU-MIDIS II 2016
Second European Union Minorities and Discrimination Survey – Main results
84
respondents in Portugal and Austria do not hold a secure status. The same applies to immigrants from South Asian countries in Greece and Italy. Only 29 % of immigrants from North Africa in Spain hold a residence permit valid for five years or longer after 10 years of residence or hold Spanish or another EU citizenship. This is an indication of the increased insecurity that long-term residents are experiencing particularly in Spain.
There are no striking differences between men and women in terms of residence status: women slightly more often hold long-term residence permits or have obtained citizenship in the country of residence. As previ-ously shown, many immigrants became citizens of their country of residence over time. Among the second gen-eration, 99 % hold a secure residence status, which to an extent is related to the fact that they are born in the EU. The next section deals with citizenship acquisition by immigrants in more detail.
Citizenship
More than half of all respondents other than Roma46 hold citizenship of their country of residence and are therefore EU citizens. Women are more often citizens (60 %) than
46 Except for Roma; Roma respondents were not asked about citizenship.
men (54 %). There are strong differences across target groups. Figure 33 shows the percentages of national citi-zens among first-generation immigrants by gender, coun-try and target group. 42 % of first-generation immigrants hold citizenship of the survey country in which they live (women: 45 %; men: 40 %). There are substantial differ-ences in the shares of citizens across target groups and EU Member States. First-generation immigrants from Turkey and North Africa in the Netherlands indicate being nation-als in the largest shares (82 %), followed by immigrants from Turkey in Sweden (75 %), immigrants from South Asia in the United Kingdom (74 %), and immigrants from Sub-Saharan Africa in Ireland (73 %). Low percentages of nationals are found among recent immigrants in Slovenia, immigrants from South Asia in Greece, and immigrants from Sub-Saharan Africa in Malta.
Strong differences between men and women are observed among Sub-Saharan African immigrants in Finland, Den-mark, Ireland and Sweden – with women indicating they are nationals at higher rates than men.
The percentage of first-generation immigrants with national citizenship increases with years of residence, which is partly related to eligibility criteria stipulated
Figure 33: National citizenship among first-generation immigrants, by gender, target group and country (%) a,b
0
40
20
80
60
10
50
30
100
90
70
AT DE DK FI FR IESSAFR TUR NOAFR (S)ASIA RIMGR Tota
lIT LU MT PT SE UK AT BE DE DK NL SE BE ES FR IT NL CY EL IT UK PL SI
7
43
5550
46
73
1512
(2)
35
50
70
50
71
26
50
82
75
59
16
41
13
82
137
74
19
(3)
42
Women Men TotalNotes: a n=8,546. b Question: “Which citizenship do you have?”Source: FRA, EU-MIDIS II 2016
85
What do the results show?
in national legislation for acquiring citizenship. When restricting the sample of first-generation immigrants to only those who were in the country for at least 10 years, the total average percentage of those who are national citizens increases to 50 %.
Among first-generation respondents with national citizen-ship, one in four (26 %) also holds a second citizenship. There are differences between Member States, influenced by their national legislation and by immigrants’ countries of origin.47 For example, there are many dual citizens among respondents from Asia in Cyprus (76 %) and among respondents from North Africa in France (64 %). Among respondents from Turkey with citizenship of the country of residence, the percentages of dual citizens range from about 1 % in Austria to 45 % in Sweden. This also reflects differences in national policies with respect to the require-ment to renounce one’s citizenship upon naturalisation.
47 As both need to allow dual citizenship.
The requirement of giving up the citizenship of one’s country of origin is the second most important reason for immigrants not to apply for naturalisation (24 %) in their country of residence; for immigrants from Turkey, this is the most important reason (51 %).
Figure 34 outlines the reasons given for not applying for citizenship by first-generation non-nationals who had never applied for citizenship. The most important reason is not fulfilling the requirements (30 %). This is strongest among recent immigrants (70 %), who obviously often do not meet the minimum residence requirements. Not fulfill-ing the requirements is not as important for respondents from Turkey compared to other target groups; as noted, for Turkish immigrants, having to give up their previous citizenship is the most important reason (51 %). Other reasons for not applying are that immigrants do not want to go through the bureaucratic procedure (17 %), or that
Figure 34: Reasons for not applying for citizenship among first-generation non-national immigrants, by target group (%) a,b,c
I do not fulfil the requirementsI do not want to give up my current citizenshipI do not want to go through the procedure (bureaucracy)I do not see any difference to my current statusI do not want itI do not think that I will stay in [COUNTRY]The fees are too expensive
0
40
20
80
60
10
50
30
100
90
70
SSAFR TUR NOAFR (S)ASIA RIMGR Total
546987
45
3
11
2026
19
51
965
813
23
13
36
999105
15
42
39
410
410
70
5711
1617
2430
5
Notes: a All immigrants without citizenship who have never applied for citizenship (n=6,649). b Question: “What are the reasons for not applying for [COUNTRY] citizenship? Any other reasons?” c The percentages add up to over 100 % because respondents could select multiple reasons.Source: FRA, EU-MIDIS II 2016
Second European Union Minorities and Discrimination Survey – Main results
86
they see no difference between their current status and citizenship (16 %). Only 11 % say that they do not want citizenship of their country of residence; 7 % think that they will not stay in the country; and 5 % state that the fees are too expensive.48
Of all non-nationals in the sample, one in five had previously applied for citizenship (20 %). Only 12 % of those who did not previously apply note that they do not want to have citizenship. This shows the major importance of citizenship for immigrants.
It is important to note that the statistics across coun-tries are weighted for the estimated size of the target population – which means, for example, that the results for the Turkish target group are mainly influenced by the results in Germany. However, due to differences in polices and migration histories, there are differences in the results across countries by target group. Regarding citizenship, the group consisting of Turkish immigrants is an interesting case, for whom the opportunities to acquire citizenship vary in the EU, but are the same due to the same country of origin.49 Figure 35 outlines the status transitions of first-generation immigrants from Turkey across the different countries; it shows that the path to citizenship is most often taken in the Netherlands, Sweden and Belgium, and that most immi-grants from Turkey in Germany hold long-term resi-dence permits (59 %). The requirement of having to give up previous citizenship is particularly relevant for this latter result.
Descendants of immigrants and citizenship status
Legal regulations on citizenship for descendants of immigrants – also referred to as the second genera-tion – vary across countries. Not all countries provide access to citizenship upon birth to children born to foreign citizens. Across the survey target groups and Member States, about one in three respondents are of the second generation (32 %). Higher shares of second- generation respondents are found among the group with Turkish background in Belgium (47 %), the Netherlands (41 %), Sweden (38 %), Germany (38 %) and Denmark (37 %). There are also higher shares of descendants of immigrants from North Africa in Bel-gium (47 %), the Netherlands (39 %) and France (38 %).
48 The percentages add up to above 100 % because multiple reasons could be selected.
49 Countries of origin can also influence to what extent migrants can take up citizenship, given that some countries do not allow expatriates to hold dual citizenship and therefore require them to renounce other citizenships. Some destination countries also require giving up second citizenships. Indicators on access to citizenship can be found, for example, on the website of the Migrant Integration Policy Index 2015.
The second generation is obviously on average much younger than the first generation (second generation: 30 years of age vs first generation: 44 years). Most sec-ond-generation respondents hold citizenship of the sur-vey country in which they live (87 %). This still means that one in 10 do not do so. Particularly among descend-ants of immigrants from Turkey, the share of national citizens is lower – at 67 %. This is mainly related to the requirement of having to give up other citizenships, which many descendants of immigrants from Turkey are not willing to do; however, they also do not become citi-zens of the country in which they reside because they do not see a difference to their current status. Looking at the indicator on secure residence status, 99 % of the second generation holds secure residence status (98 % among second-generation respondents with Turkish background). The share of dual citizens among second-generation respondents is 11 % overall, and particularly high for second-generation respondents with Turkish background in Sweden (54 %), Belgium (36 %), the Netherlands (32 %) and Denmark (29 %).
Family reunification
Among respondents who are immigrants or descend-ants of immigrants, 7 % have a wife, husband or chil-dren living outside the country in which they live. The percentage of those with family outside the coun-try is higher among groups with recent immigration histories, particularly among immigrants from South Asia in Greece (44 %) and immigrants from Asia in Cyprus (29 %). Among recent immigrants in Slovenia, one in five has family outside the country (21 %). The same rate can be found among immigrants from Sub-Saharan Africa in several countries, including Italy, Lux-embourg, Malta and Austria (19–20 %). For immigrants and descendants of immigrants from Turkey and North Africa, the percentage of respondents with close family outside the country is below 10 % in all countries but Italy (13 % of respondents with North African back-ground). This indicates a lower potential family reuni-fication rate for some groups. The percentage of those with family outside the country is somewhat higher among men than women (overall, 8 % for men and 6 % for women).
Among those with family outside the country, one in five have previously applied for family reunification. A higher number of respondents with family outside the country already applied for family reunification in Finland: of those with family outside the country (15 % of all respondents there), almost every second (47 %) has previously applied for family reunification. Of all respondents who previously applied for family reunifi-cation, slightly more than one third received a positive decision (36 %); the applications by a third of respond-ents were rejected (33 %). The remaining respondents
87
What do the results show?
Figure 35: Status upon arrival and current residence status for first-generation immigrants from Turkey by Member State (%) a,b,c
(4)
(0)
(4)
9
(0)
82
(3)
(5)
(3)
(0)
21
64
(5)
NL
(10)
(2)
(1)
11
(1)
75
(5)
(7)
(1)
(1)
21
23
26
17
SE
5
(1)
4
(0)
59
4
26
(2)
4
4
(1)
2
37
43
6
DE
(3)
(4)
42
(1)
50
(1)
(5)
12
(1)
(1)
24
54
(2)
DK
(6)
(3)
14
(4)
(2)
71
(3)
(2)
6
(0)
(2)
28
43
16
BE
(2)
(2)
19
(0)
27
(0)
50
(2)
(1)
23
6
(2)
21
41
(4)
AT
Status upon arrival Current status
Status upon arrival Current status
Student visa
Other
No permit
Tourist visa
Asylum seeker
Parents' papers
A residence/work permit
No need for papers
Other
No permit
Tourist visa
Student visa
Asylum seeker
Parents' papers
A residence/work permit
No need for papers
No permit
Other
Tourist visa
Student visa
Asylum seeker
Parents' papers
A residence/work permit
No need for papers
Student visa
Other
No permit
Tourist visa
Asylum seeker
Parents' papers
A residence/work permit
No need for papers
Asylum seeker
Other
No permit
Tourist visa
Parents' papers
A residence/work permit
No need for papers
Other
No permit
Tourist visa
Student visa
Asylum seeker
Parents' papers
A residence/work permit
No need for papers
Other
No residence permit
Limited permit < 5 years
Limited permit > 5 years
Unlimited residence permit
EU citizen
National citizen
Other
No residence permit
Limited permit < 5 years
Unlimited residence permit
EU citizen
National citizen
Other
No residence permit
Limited permit < 5 years
Limited permit > 5 years
Unlimited residence permit
EU citizen
National citizen
Other
Limited permit < 5 years
Unlimited residence permit
EU citizen
National citizen
Other
No residence permit
Limited permit < 5 years
Unlimited residence permit
EU citizen
National citizen
Other
No residence permit
Limited permit < 5 years
Unlimited residence permit
EU citizen
National citizen
Second European Union Minorities and Discrimination Survey – Main results
88
Figure 35 (continued)
(4)
(0)
(4)
9
(0)
82
(3)
(5)
(3)
(0)
21
64
(5)
NL
(10)
(2)
(1)
11
(1)
75
(5)
(7)
(1)
(1)
21
23
26
17
SE
5
(1)
4
(0)
59
4
26
(2)
4
4
(1)
2
37
43
6
DE
(3)
(4)
42
(1)
50
(1)
(5)
12
(1)
(1)
24
54
(2)
DK
(6)
(3)
14
(4)
(2)
71
(3)
(2)
6
(0)
(2)
28
43
16
BE
(2)
(2)
19
(0)
27
(0)
50
(2)
(1)
23
6
(2)
21
41
(4)
AT
Status upon arrival Current status
Status upon arrival Current status
Student visa
Other
No permit
Tourist visa
Asylum seeker
Parents' papers
A residence/work permit
No need for papers
Other
No permit
Tourist visa
Student visa
Asylum seeker
Parents' papers
A residence/work permit
No need for papers
No permit
Other
Tourist visa
Student visa
Asylum seeker
Parents' papers
A residence/work permit
No need for papers
Student visa
Other
No permit
Tourist visa
Asylum seeker
Parents' papers
A residence/work permit
No need for papers
Asylum seeker
Other
No permit
Tourist visa
Parents' papers
A residence/work permit
No need for papers
Other
No permit
Tourist visa
Student visa
Asylum seeker
Parents' papers
A residence/work permit
No need for papers
Other
No residence permit
Limited permit < 5 years
Limited permit > 5 years
Unlimited residence permit
EU citizen
National citizen
Other
No residence permit
Limited permit < 5 years
Unlimited residence permit
EU citizen
National citizen
Other
No residence permit
Limited permit < 5 years
Limited permit > 5 years
Unlimited residence permit
EU citizen
National citizen
Other
Limited permit < 5 years
Unlimited residence permit
EU citizen
National citizen
Other
No residence permit
Limited permit < 5 years
Unlimited residence permit
EU citizen
National citizen
Other
No residence permit
Limited permit < 5 years
Unlimited residence permit
EU citizen
National citizen
Notes : a Out of all respondents who are first-generation immigrants from Turkey (upon arrival: n=2,067; current status: n=2,127); weighted results.
b Based on questions: “Did you have any type of permit, documents or status when you first arrived in [COUNTRY]? If yes, which one from the following list did you have?”; “Do you currently have a permit to stay in [COUNTRY]?”; “Altogether, for how long is your current permit valid?”
c Thickness of lines indicates the share of respondents reporting the different residence status upon arrival and current status. It shows the flow from arrival status to current status.
Source: FRA, EU-MIDIS II 2016
89
What do the results show?
are still awaiting a decision (28 %) or have withdrawn their application (3 %).
Among respondents with family outside the country who have not previously applied for family reunifica-tion, over a third indicate that their spouses or children do not want to come (19 %) or that they themselves do not want, or there is no need, to apply for family reunification (18 %).
Another third of those who did not apply for family reunification do not fulfil the requirements (18 %), could not afford to sustain their family members if they joined them (6 %), do not want to go through the procedure (5 %), or find the fees too expensive (4 %).
The remaining respondents either indicated that they did not know the answer to the question (5 %), planned to leave the country (3 %), intended to divorce (1 %) or had other reasons (22 %).50
Not meeting the requirements is especially impor-tant among immigrants from Sub-Saharan Africa in Malta (41 %), France (31 %) and Austria (29 %), as well as among immigrants from South Asia in Greece, who also often indicate that the fees are too expensive (29 %). For the latter group, not being able to sustain the family is even more of an issue (41 %). This is also the reason given for not applying in Italy, particularly by immigrants and descendants of immigrants from North Africa (40 %) and Sub-Saharan Africa (36 %). Immigrants from Asia in Cyprus most often have not applied because of plans to return to their home coun-tries (27 %) or because their family does not want to come (25 %).
2�4�2� Education and language proficiency
The right to education is protected under Article 28 of the UN Convention on the Rights of the Child – ratified by all EU Member States – and Article 14 of the EU Charter of Fundamental Rights. EU Member States are obliged to ensure that all children enjoy equal access to edu-cation, particularly compulsory education. According to UNESCO, individuals who do not complete at least compulsory education face a high risk of living in pov-erty and have limited opportunities to develop learning skills and reach their full potential.51
Level of educational attainment
Education is vital for better chances in the labour mar-ket. The EU considers completing upper secondary
50 The numbers add up to over 100 % due to rounding.51 UNESCO (2010), p. 155.
education an important step not just for successful entry into the labour market, but also as a basic requirement for participating in the kind of further training that is necessary to succeed in a knowledge-based society.52 In light of this, EU-MIDIS II asked about the highest educa-tion level completed either in the country of residence or in any other country inside or outside of the EU. Only 61 % of EU-MIDIS II respondents aged 16-64 years reported having completed at least upper secondary education, with significant variations across countries as well as target groups. By contrast, in the EU-28, on aver-age, 74 % of the population aged 15 to 64 years53 had completed at least upper secondary education in 2016.54
As already presented in FRA’s 2016 report on EU-MIDIS II findings on Roma,55 EU-MIDIS II results indicate persist-ing low levels of educational achievement among the Roma population. In seven of the nine countries in which Roma were surveyed, the percentage of Roma aged 16 to 64 who have completed upper secondary education remains below 15 %. The proportion is higher only in the Czech Republic (34 %) and Slovakia (32 %), but still far from the EU average. No other target group covered in EU-MIDIS II indicates such a low educational level over-all. Low levels of education – with no upper secondary education received – are also observed among respond-ents with Turkish background in Germany (42 %); with Sub-Saharan African background in Malta (11 %), Por-tugal (33 %) and Italy (42 %); with North African back-ground in Italy (30 %) and Spain (43 %); and with South Asian background in Italy (29 %) and Greece (43 %) (Figure 36). On the opposite end, recent immigrants in Poland (95 %), respondents with Sub-Saharan African background in Ireland (83 %) and respondents belong-ing to the Russian minority in Estonia (88 %) and Lat-via (90 %) exceed the educational level of the general population in these countries.
The educational level of second-generation respond-ents is on average higher than that of first-generation respondents in all countries – with the exception of Fin-land, where only 68 % of second-generation respond-ents with Sub-Saharan African background completed at least upper secondary education, compared with 83 % of first-generation respondents and well below the gen-eral population (81 %). In France, Portugal and Sweden, the shares of second-generation respondents with at least upper secondary education (89 %, 56 %, 87 %, respectively) even exceed the shares reported for the general population living in those countries (75 %, 47 %
52 Council Conclusions (2003/C 134/02) of 5 May 2003 on reference levels of European average performance in education and training (Benchmarks), 2003.
53 EU-MIDIS II only allows calculating the values for the age group 16-64 because respondents had to be at least 16 years old. See Annex I on Methodology.
54 Eurostat edat_lfse_03 (download 11/07/2017); age group 15–64 years; Labour Force Survey.
55 FRA (2016), p. 47.
Second European Union Minorities and Discrimination Survey – Main results
90
and 79 % respectively).56 Improvements in educational level of second-generation vis-à-vis first-generation respondents vary across target groups and countries. For example, among respondents with Turkish back-ground, the difference between the first and second
56 In Cyprus, Greece and Ireland, the EU-MIDIS II sample did not contain a sufficient number of respondents belonging to the second generation.
generations in Austria – in terms of having completed at least upper secondary education – is 41 percentage points (36 % for first- and 77 % for second-genera-tion respondents). In Germany, almost no difference between first- and second-generation respondents with
Figure 36: Respondents aged 16–64 years who have attained upper secondary, post-secondary non-tertiary or tertiary education (ISCED 2011 levels 3–8) in any country, compared with the general population (Eurostat 2016), by target group and EU Member State (%) a,b,c,d,e,f
0
40
20
80
60
10
50
30
100
90
70
ATSSAFR RIMGR
DK FR IT MT SEDE FI IE LU PT UK PL SI
40 45 48 76 7550 66 64 78 91(11) 11
30 36 64 69 58 66 95 96 68 8584 35 4675 63 66
80
48
80
56
73
59
81
80 75
757583
58
42
72
65
49
11
47
33
79
67
8580
95
62
83
77
Women General population (Eurostat 2016)Men Total
0
40
20
80
60
10
50
30
100
90
70
TUR NOAFR (S)ASIAAT SE ITBE NLDE BEDK ES CY ITNL FR EL UK
50 50 52 54 40 44 60 57 57 57 80 77 61 66 42 44 81 74 32 28 73 61 54 66 42 22 33 60 59
80
50
72
53
80
42
73
59
72
57
79
78 72
64
72
67
57
43
75
78
58
30
76
5958
29
70
43
80
60
91
What do the results show?
Figure 36 (continued)
Women General population (Eurostat 2016)Men Total
0
40
20
80
60
10
50
30
100
90
70
RUSMINBG
ROMAEL HR PT SK EECZ ES HU RO LT LV
34 34
13 16
28 37 94 78 82 81 92 88
6 7
78
1
88
34
70
2
57
3
80
12
78
47
0
72
6
86
32
85
90
83
88 88
81
14
Notes: a Out of all respondents aged 16–64 years (n=23,610); weighted results. b Highest educational level attained either in the country, where a respondent was interviewed or in any other country. c General population 2016: Eurostat edat_lfse_03 (download 11/07/2017); age group 15–64 years; Labour Force Survey. d Results based on a small number of responses are statistically less reliable. Thus, results based on 20 to 49 unweighted
observations in a group total or based on cells with fewer than 20 unweighted observations are noted in parentheses (results by gender not presented for Roma in some countries due to less than 20 observations in each). Results based on fewer than 20 unweighted observations in a group total are not published.
e Acronyms for target groups refer to immigrants from [country/region] and their descendants: TUR = Turkey, SSAFR = Sub-Saharan Africa, NOAFR = North Africa, (S)ASIA = South Asia and Asia, RIMGR = recent immigrants from non-EU countries, RUSMIN = Russian minority, ROMA = Roma minority.
f Question: “What is the highest level of education you have completed?”Source: FRA, EU-MIDIS II 2016; Eurostat database
Turkish background was observed (40 % and 44 %, respectively) – by far the smallest difference among the countries in which this target group was interviewed. Although EU-MIDIS II does not provide information on intergenerational mobility, these results indicate such an effect.
Among respondents who completed at most lower secondary education (ISCED 2011 levels 0-2), on aver-age, 38 % of those aged 16-64 were no longer pursu-ing any form of education or training at the time of EU-MIDIS II – even though attaining the higher educa-tion would increase their chances in the labour market. The situation varies significantly across target groups and Member States. Almost all Roma respondents in this age group in Greece, Portugal, Bulgaria, Spain and Romania (98 % – 92 %) are no longer pursuing edu-cation and have completed at most lower secondary education. The values are similar for respondents with Sub-Saharan African background in Malta (84 %), fol-lowed by respondents with North African and South
Asian backgrounds in Italy (67 % and 66 %). The share of respondents with low levels of education who are not pursuing further education or training decreases for respondents of younger age (Figure 37) and for second-generation respondents. On average, only every fifth respondent aged 16 to 24 who has completed at most lower secondary education is not continuing education or training. The same applies for second-generation respondents.
Language proficiency
Learning the language(s) of one’s country of residence is an important aspect of integrating into society. EU-MIDIS II respondents were asked to assess their proficiency in the national language of their country of residence in terms of speaking, reading and writing. Where more than one national language is used in a country, respondents were asked about their proficiency in at most two national lan-guages. They could choose from the following scale: no skills, not good skills at all, not so good, good, excellent
Second European Union Minorities and Discrimination Survey – Main results
92
and mother tongue. For further analysis, the lower three categories were merged into one category – “insufficient language proficiency” – and the upper three into another category – “sufficient language proficiency”.
On average, 71 % of respondents indicated having sufficient proficiency in their country of residence’s national language(s) in terms of speaking, reading and writing. As many as 91 % of respondents with Sub-Saharan African background indicate having such proficiency in Ireland, and 90 % of this target group do so in the United Kingdom. High shares are also indi-cated by respondents with Sub-Saharan African and North African backgrounds in France (85 % and 82 %, respectively), and respondents with North African background in Belgium (81 %). High language profi-ciency levels for these target groups and countries are not surprising, given the colonial histories of the United Kingdom, France and Belgium (French language). In comparison, respondents with Turkish background indicate having sufficient language proficiency at
lower rates (60% – 75%) – with the exception of Swe-den, where 83 % of respondents from this group report having ‘good’ to mother-tongue-level proficiency in Swedish (Figure 38).
Nearly all second-generation respondents, irrespective of their parents’ country of origin or residence, indicated having ‘good to mother tongue’ language proficiency in terms of speaking, reading and writing (Table 2).
Members of the Russian minority and Roma face a different situation with respect to their competency in their countries’ official languages. Despite often being mother tongue speakers of the national lan-guage, Roma still indicate low education levels. The Russian minority is a recognised minority in the Baltic states and particularly elderly people are less fluent in the national languages. Table 3 gives an overview of the language competencies of Roma and Russian minority respondents in the three areas of speaking, writing and reading.
Figure 37: Respondents aged 16–64 years who have completed at most lower secondary education (ISCED 2011 levels 0-2) in any country and do not continue with further education or training, by target group and age (%) a,b,c,d,e
0
40
20
80
60
10
50
30
100
90
70
SSAFR TUR NOAFR (S)ASIA ROMA RUSMINRIMGR Total
30
40
6
48 48
59
15
81
26
37
7
71
40
55
10
63
13
18
(5)
81
88
72
96
6
38
48
17
66
16-24 25-44 45-59 60+ Total
29
44
27
39
12
82
8
38
10(8) 11
Notes: a Out of all respondents aged 16–64 years (n=23,571); weighted results. b Highest educational level attained either in the country where a respondent was interviewed or in any other country. c Results based on a small number of responses are statistically less reliable. Thus, results based on 20 to 49 unweighted
observations in a group total or based on cells with fewer than 20 unweighted observations are noted in parentheses. Results based on fewer than 20 unweighted observations in a group total are not published.
d Acronyms for target groups refer to immigrants from [country/region] and their descendants: TUR = Turkey, SSAFR = Sub-Saharan Africa, NOAFR = North Africa, (S)ASIA = South Asia and Asia, RIMGR = recent immigrants from non-EU countries, RUSMIN = Russian minority, ROMA = Roma minority.
e Questions: “What is your main activity status?”; “What is the highest level of education you have completed?”; “Do you currently attend school or vocational training?”
Source: FRA, EU-MIDIS II 2016
93
What do the results show?
Figure 38: Respondents with ‘good to mother tongue’ language proficiency (in all three dimensions – speaking, reading and writing) of at least one national language in the country where interviewed, by target group and EU Member State (%) a,b,c,d,e
0
40
20
80
60
10
50
30
100
90
70
2925
6059
6865
52
69
8684
96
87
4635
7976
36
(10)
7569
55
70
8990
6061
7773
6771
6167
7369
8187 8280
5953
76
88
4945
80
65
28
60
67
59
8591
41
78
35
72
61
90
60
7569
64
71
83 81
56
82
47
73
AT DE DK FI FR IESSAFR TUR NOAFR
IT LU MT PT SE UK AT BE DE DK NL SE BE ES FR IT NL
Women Men Total
0
40
20
80
60
10
50
30
100
90
70
26
49
13
43
29
82
71
38
62
4950
59
50
6567
3833
78
7172
60
6769
53
29
4242 4444
(16)
26
31
43
3339
40
14
37
76
51 4954
66
35
74
66 68
40 42 44
22
38 36
Women Men Total(S)ASIA RIMGR
CY EL IT UK PL SIRUSMIN
BGROMA
EL HR PT SK EECZ ES HU RO LT LV
Notes: a Out of all respondents (n=25,412); weighted results. b Good, excellent and mother tongue level proficiency in all three dimensions (speaking, reading and writing) of at
least one national language in the country where the respondent was interviewed. c Results based on a small number of responses are statistically less reliable. Thus, results based on 20 to
49 unweighted observations in a group total or based on cells with fewer than 20 unweighted observations are noted in parentheses. Results based on fewer than 20 unweighted observations in a group total are not published (Female (S)ASIA EL).
d Acronyms for target groups refer to immigrants from [country/region] and their descendants: TUR = Turkey, SSAFR = Sub-Saharan Africa, NOAFR = North Africa, (S)ASIA = South Asia and Asia, RIMGR = recent immigrants from non-EU countries, RUSMIN = Russian minority, ROMA = Roma minority.
e Question: “Using this scale, how would you describe your proficiency in [SURVEY COUNTRY NATIONAL LANGUAGE 1/2] as regards speaking/reading/writing?”
Source: FRA, EU-MIDIS II 2016
Second European Union Minorities and Discrimination Survey – Main results
94
Table 2: Language proficiency of respondents (speaking, reading and writing) in at least one national language in the country of residence, by target group and generation (%) a,b,c,d,e
Good to mother tongue language proficiency in all dimensions (speaking, reading and writing)
Weak to no language proficiency in all three dimensions (speaking, reading and writing)
First Generation
Second Generation
Total First Generation
Second Generation
Total
SSAFR 77 98 82 8 (1) 7TUR 53 95 69 24 (1) 15NOAFR 68 96 78 13 (1) 9(S)ASIA 62 97 70 20 (1) 15
Notes: a Out of all respondents in listed target groups (n=15,282); weighted results. b Good, excellent and mother tongue level proficiency in all three dimensions (speaking, reading and writing) of at least
one national language in the country where respondent was interviewed. No skills, not good at all and not so good proficiency in all three dimensions (speaking, reading and writing) of at least one national language in the country where respondent was interviewed.
c Results based on a small number of responses are statistically less reliable. Thus, results based on 20 to 49 unweighted observations in a group total or based on cells with fewer than 20 unweighted observations are noted in parentheses. Results based on fewer than 20 unweighted observations in a group total are not published.
d Acronyms for target groups refer to immigrants from [country/region] and their descendants: TUR = Turkey, SSAFR = Sub-Saharan Africa, NOAFR = North Africa, (S)ASIA = South Asia and Asia.
e Question: “Using this scale, how would you describe your proficiency in [SURVEY COUNTRY NATIONAL LANGUAGE 1/2] as regards speaking/reading/writing?”
Source: FRA, EU-MIDIS II 2016
Table 3: Language proficiency with regard to speaking, reading and writing the national language in their country, for Roma and Russian minority (%) a,b,c,d
Speaking Writing ReadingWeak to no skills
Good to mother tongue
Weak to no skills
Good to mother tongue
Weak to no skills
Good to mother tongue
ROMA BG 9 91 45 55 38 62CZ 8 92 34 66 23 77EL 5 95 64 36 56 44ES 2 98 26 74 21 79HR 13 87 32 68 30 70HU 2 98 30 70 29 71PT 9 91 60 40 54 46RO 14 86 56 44 50 50SK 26 74 55 45 44 56Total 12 88 45 55 38 62
RUSMIN EE 70 30 74 26 69 31LT 32 68 61 39 43 57LV 57 43 60 40 55 45Total 58 42 65 35 59 41
Notes: a Out of all respondents (n=4,256); weighted results. b Good to mother tongue proficiency answer categories: good, excellent and mother tongue. Weak to no skills answer
categories: not so good, not good at all, no skills. In at least one national language in the country where respondent was interviewed.
c Acronyms for target groups: RUSMIN = Russian minority, ROMA = Roma minority. d Question: “Using this scale, how would you describe your proficiency in [SURVEY COUNTRY NATIONAL LANGUAGE 1/2]
as regards speaking/reading/writing?”Source: FRA, EU-MIDIS II 2016
95
What do the results show?
Immigrants’ and descendants of immigrants’ compe-tencies in speaking, writing and reading the national language of their country of residence increase with their education levels. On average, 85 % of respondents with at least upper secondary education have sufficient proficiency of the national language in all three dimen-sions. By contrast, only 7 % of respondents with this education level are insufficiently proficient in all three aspects of the respective language – these may consist of immigrants who complete their higher education before coming to their current country of residence (Figure 39).
The overall ability to speak, read and write in a country’s language provides a strong basis for integration into society, including the labour market. The ability to read in the national language provides access to both rights and obligations. The figures on language proficiency presented above do provide hope by indicating high
shares of people with sufficient language proficiency. However, on average, 9 % of respondents still do not have sufficient proficiency of the national language when it comes to reading. As many as 4 % of respond-ents cannot read in their country of residence’s national language(s) at all. This means that they have no chance to learn about their rights should information only be available in the national language.
2�4�3� Labour market participation
Employment is vital for individuals’ societal integration. Article 15 of the EU Charter of Fundamental Rights pro-tects the right to engage in work. Employment is also the first headline indicator of the Europe 2020 Strategy– aiming at 75 % of the EU population aged 20 to 64 being employed by 2020.
Figure 39: Respondents with ‘good to mother tongue’ language proficiency in the country of residence, by level of education (ISCED) and target group (%) a,b,c,d,e,f
0
40
20
80
60
10
50
30
100
90
70
SSAFR TUR NOAFR (S)ASIA RIMGR Total
91
65
83
59
93
46
87
52 52
40
75
51
At most lower secondary education At least upper secondary education
Notes: a Out of all respondents (n=25,347); weighted results. b ‘Good to mother tongue’ – good, excellent and mother tongue level proficiency in all three dimensions (speaking,
reading and writing) of at least one national language in the country where respondent was interviewed. c At most lower secondary education (less than primary or primary and lower secondary education); at least upper
secondary education (upper secondary, or post-secondary non-tertiary or tertiary education). d Acronyms for target groups refer to immigrants from [country/region] and their descendants: TUR = Turkey,
SSAFR = Sub-Saharan Africa, NOAFR = North Africa, (S)ASIA = South Asia and Asia, RIMGR = recent immigrants from non-EU countries
e Question: “Using this scale, how would you describe your proficiency in [SURVEY COUNTRY NATIONAL LANGUAGE 1/2] as regards speaking/reading/writing?”
f Remainder to 100 are respondents with a given level of education, but with some deficiencies in language proficiency in speaking, reading or writing.
Source: FRA, EU-MIDIS II 2016
Second European Union Minorities and Discrimination Survey – Main results
96
EU-MIDIS II data allow for calculation of the paid work rate, a rough approximation of the employment rate used by Eurostat to measure Europe 2020 targets.57 ‘Paid work’ here refers to household members who declared their main activity as being ‘employed’ or ‘self-employed’ at the time of the survey,58 including those who did some work in the previous four weeks to earn some money (Figure 40).
On average, 61 % of household members of working age reported that they were employed or had done some work in the four weeks preceding the survey. This is below the average EU-28 employment rate of 71 % in 2016. EU-MIDIS II data indicate that members of the Russian minority in the Baltic countries have about the same employment rates as the general population.
57 The calculated paid work rate is not fully comparable to the ILO concept based on the Eurostat employment rate, which defines as employed persons 15 years or older who have worked for at least one hour for pay or profit or family gain during the reference week or persons who were not at work during the reference week but had a job or business from which they were temporarily absent. The paid work rate as calculated in EU MIDIS II is based on the respondent questionnaire on self-declared current main activity. If the main activity was indicated as ‘inactive’ or ‘unpaid’, the person was asked if they “did any work in the last four weeks to earn some money”. This question intended to also capture informal work and small jobs that may contribute to a family’s survival. In contrast to the ILO concept, unpaid help and parental leave are not explicitly included in the definition of ‘paid work’.
58 Unpaid helpers in family businesses were excluded from this definition.
The largest difference in employment rate from the general population was observed for Roma (by up to 40 percentage points in Croatia and Spain). Respond-ents with Sub-Saharan African background in Portugal (77 %), with North African background in Italy (70 %), with Asian background in Cyprus (83 %), and with South Asian background in Greece (82 %) and Italy (67 %), as well as recent immigrants in Poland (83 %) and Slove-nia (76 %), indicate engaging in paid work at higher rates than the employment rate reported for the gen-eral population in these countries.
In almost all target groups and countries ― with the exception of respondents with Sub-Saharan African background in Austria and Portugal, with Turkish back-ground in Sweden, with Asian background in Cyprus,
Figure 40: Paid work rate for household members aged 20–64 years, including self-employment and occasional work or work in the past 4 weeks, compared with the Europe 2020 employment rate 2015 (Eurostat), by target group and EU Member State (%) a,b,c,d,e
0
40
20
80
60
10
50
30
100
90
70
ATSSAFR RIMGR
DK FR IT MT SEDE FI IE LU PT UK PL SI
Women General population employment rate (Eurostat 2016)Men Total
60 68 36 62 6340 71 49 44
66
50
6367
49
83
76
63
75
68
44
69
70
62
71 70
8178
6971 70
7073
777975
58 (30) 51 80 72 59 66 74 77 78 88 52 9272 43 7370 65 68
77
45
97
What do the results show?
Figure 40 (continued)
0
40
20
80
60
10
50
30
100
90
70
TUR NOAFR (S)ASIAAT SE ITBE NLDE BEDK ES CY ITNL FR EL UK
30 81 35 64 54 80 54 66 43 62 70 70 37 58 24 51 50 69 46 86 42 64 84 81 (21) 85 23 93 32 71
5550
68
60
5348
38
59
70
83 82
67
5552
70
75
68
79 77 7781
6864
70
62
69
5662
77 78
Women General population employment rate (Eurostat 2016)Men Total
0
40
20
80
60
10
50
30
100
90
70
RUSMINBG
ROMAEL HR PT SK EECZ ES HU RO LT LV
35 64 32 5522
82 36 62 32 54 75 76 73 79 73 8316 31
21 55
12 31
27 64
52
24
49
38
7675
4643
78
21
4349
56
64 61
72 7166
70
77
75 73
77
68
Notes: a Out of all household members aged 20–64 years (n=44,877); weighted results. b Europe 2020 employment rate 2016: Eurostat t2020_10 (download 11/07/2017). This is calculated by dividing the
number of persons aged 20 to 64 in employment by the total population of the same age group. The indicator is based on the ILO concept, Labour Force Survey.
c Results based on a small number of responses are statistically less reliable. Thus, results based on 20 to 49 unweighted observations in a group total or based on cells with fewer than 20 unweighted observations are noted in parentheses. Results based on fewer than 20 unweighted observations in a group total are not published.
d Acronyms for target groups refer to immigrants from [country/region] and their descendants: TUR = Turkey, SSAFR = Sub-Saharan Africa, NOAFR = North Africa, (S)ASIA = South Asia and Asia, RIMGR = recent immigrants from non-EU countries, RUSMIN = Russian minority, ROMA = Roma minority.
e Questions: “Please look at this card and tell me which of these categories describes your current situation best?”; “Did you do any work in the last 4 weeks to earn some money?”
Source: FRA, EU-MIDIS II 2016
Second European Union Minorities and Discrimination Survey – Main results
98
and of the Russian minority in Estonia ― a large gender gap can be observed, with fewer women engaged in paid work. For respondents with Turkish, North Afri-can, and (South) Asian backgrounds, as well as for Roma respondents and those of the Russian minority, the paid work rate increases with age, and is highest among those aged 35 to 44. Respondents with Sub-Saharan African background, as well as recent immi-grants in Poland and Slovenia, reported the highest paid work rates for the group aged 45 to 54. Across all target groups, no differences in paid work rates were observed between first-generation respondents and second-generation respondents.
The data confirm that a higher educational level gen-erally corresponds to a better labour market position. On average, only 52 % of respondents who indicated having completed at most lower secondary education reported being in paid work, compared to 67 % of those who completed at least upper secondary education. Among individual target groups, completing more than just lower secondary education has the largest influ-ence on labour market participation for Roma respond-ents, followed by respondents with North African background (Figure 41).
When looking at the different target groups in specific Member States, labour market participation is most strongly connected to education levels for respond-ents with North African background in the Netherlands, with a paid work rate of 28 % for those with at most lower secondary education, and of 60 % for those with at least upper secondary education; respondents with Sub-Saharan African background in Denmark (rates of 27 % and 50 %, respectively) and Sweden (37 % and 67 %); and respondents with Turkish background in Denmark (38 % and 72 %) and the Netherlands (35 % and 65 %). Obtaining educational levels higher than the lower secondary level also increases the paid work rate by at least 100 % for Roma in the Czech Republic.
Similarly to education, the ability to participate in the labour market is also related to national language profi-ciency. On average, only 43 % of those with insufficient language proficiency in all three dimensions (speak-ing, reading and writing) reported being in paid work in the four weeks before the survey (Table 4). Among respondents with sufficient language proficiency in all three dimensions, 65 % reported being in paid work.
Being engaged in paid work is most strongly related to language proficiency for respondents with Turkish background. With the exception of such respondents
Figure 41: Paid work rate for respondents aged 20–64 years, by education level and target group (%) a,b,c,d
0
40
20
80
60
10
50
30
100
90
70
SSAFR TUR NOAFR (S)ASIA RUSMINROMARIMGR Total
4446
77
62 60
43
64
46
80
54
61
40
76
6367
52
ISCED levels 0-2 - less than primary, primary and lower secondary education
ISCED levels 3-8 - upper secondary, post-secondary non-tertiary and tertiary education
Notes: a Out of all respondents aged 20–64 years (n=21,578); weighted results. b Highest educational level attained either in the country where a respondent was interviewed or in any other country. c Acronyms for target groups refer to immigrants from [country/region] and their descendants: TUR = Turkey,
SSAFR = Sub-Saharan Africa, NOAFR = North Africa, (S)ASIA = South Asia and Asia, RIMGR = recent immigrants from non-EU countries, RUSMIN = Russian minority, ROMA = Roma minority.
d Question: “Please look at this card and tell me which of these categories describes your current situation best?”Source: FRA, EU-MIDIS II 2016
99
What do the results show?
in Austria, sufficient language proficiency increases the paid work rate for this group by more than 100 % ― on average, from 35 % to 78 %. A similarly high effect was observed for respondents with Asian background in Italy, for Sub-Saharan Africans in Austria and Malta, North Africans in Spain, and Roma in Slovakia. The connection between language proficiency and engagement in paid work seems to be even stronger for respondents with Sub-Saharan African background in Italy (72 % of those with sufficient language proficiency in speaking, reading and writing in paid work, compared to 22 % of those with insufficient language proficiency) and in Sweden (71 % vs 25 %). The correlation between language proficiency and engagement in paid work is weakest for recent immi-grants in Poland and Slovenia. For this group in these two countries, similarly high paid work rates were observed regardless of language proficiency ― with values around the same as the general population’s employment rate in the two countries.
2�4�4� Trust in public institutions and political participation
The level of trust in different institutions varies consider-ably across countries and target groups. Across all tar-get groups and countries, respondents show the highest level of trust in the local (municipal) authorities and in the police ― with an average value of 6.3 on a scale from 0 to 10, where 0 means ‘no trust at all’ and 10 represents ‘complete trust’. This is closely followed by trust in the
legal system, at 6.1. The general level of trust in coun-tries’ parliaments lies at 5.3, and is 4.9 for the European Parliament. Lower levels of trust are indicated towards countries’ politicians (4.2) and political parties (4.1). Trust in local authorities is highest in most countries and target groups ― specifically, in 21 of the 42 country and target group combinations. This is followed by trust in the police ― in 15 of the 42 country and target group combinations. Trust levels are lowest for either politicians or political parties in most combinations (36 out of 42).
These results can be compared to those for the gen-eral population based on data from the 2014 European Social Survey (ESS), which asked the same questions in 18 EU Member States.59 The police (6.2) and the legal system (5.1) are also the most highly ranked authori-ties/institutions in terms of trust among the general population. (The ESS 2014 did not specifically ask about trust in local authorities.) In the ESS 2014, the general population overall indicates the highest levels of trust in the police, with results highest in Finland (7.9) and Denmark (7.6) and lowest in Poland (5.1). In most of the 18 Member States (14 out of 18), either political parties or politicians are ranked lowest. In all of these countries, the average trust level is below 5 for politicians, and is lowest in Slovenia at 1.9.
59 The ESS 2014 provides data on the following countries: Austria, Belgium, the Czech Republic, Denmark, Estonia, Finland, France, Germany, Hungary, Ireland, Lithuania, the Netherlands, Poland, Portugal, Slovenia, Spain, Sweden and the United Kingdom.
Table 4: Paid work rate for respondents aged 20–64 with sufficient and insufficient language proficiency (in all three dimensions – speaking, reading and writing) of at least one national language in the country where interviewed, by target group (%) a,b,c,d,e
Insufficient language proficiency in all dimensions (speaking, reading and writing)
Sufficient language proficiency in all dimensions (speaking,
reading and writing)SSAFR 53 72TUR 35 78NOAFR 36 56(S)ASIA 30 64RIMGR 71 81ROMA 36 46RUSMIN 70 82Total 43 65
Notes: a Out of all respondents aged 20–64 years in listed target groups (n=21,536); weighted results. b Sufficient proficiency – good, excellent and mother tongue level proficiency in all three dimensions (speaking, reading
and writing) of at least one national language in the country where respondent was interviewed. c Insufficient proficiency – no skills, not good at all and not so good proficiency in all three dimensions (speaking,
reading and writing) of at least one national language in the country where respondent was interviewed. d Acronyms for target groups refer to immigrants from [country/region] and their descendants: TUR = Turkey,
SSAFR = Sub-Saharan Africa, NOAFR = North Africa, (S)ASIA = South Asia and Asia. e Question: “Using this scale, how would you describe your proficiency in [SURVEY COUNTRY NATIONAL LANGUAGE 1/2]
as regards speaking/reading/writing?”Source: FRA, EU-MIDIS II 2016
Second European Union Minorities and Discrimination Survey – Main results
100
A closer look at levels of trust in the police and the legal system ― as important areas with higher levels of trust ― follows, including comparisons with the ESS.
Figure 42 shows the levels of trust in the police and Figure 43 shows the levels of trust in the legal sys-tem by country and target group in EU-MIDIS II, compared with the general population in the same countries, where available.
Immigrants and descendants of immigrants from Turkey indicate higher levels of trust than most other EU-MIDIS II target groups, with results similar to those for the general population. Levels of trust in the police are the same or higher than the general population’s in Belgium, Austria and Germany, and slightly lower in Denmark, Sweden and the Netherlands. Levels of trust in the legal system are very similar to, or slightly above, those of the general population in the countries in which immigrants and descendants of immigrant from Tur-key were surveyed ― apart from Denmark, where the average trust level is somewhat lower. Immigrants and descendants of immigrants from South Asia also show high levels of trust in the police and the legal system, which is influenced by the high levels of this target group in in the United Kingdom.
Among immigrants and descendants of immigrants from Sub-Saharan Africa, levels of trust in the police are about the same or slightly higher than the general population’s in most countries ― but not in Sweden, Denmark and Austria. In the latter, trust in the police is at 3.6 (meaning there is a general tendency to distrust the police). Trust in the legal system is higher among immigrants and descendants of immigrants from Sub-Saharan Africa than among the general population in most countries covered, with stronger differences in
Portugal, Germany, Ireland and Finland. Only in Austria and Denmark are the levels of trust in the legal system clearly lower than among the general population. There is no benchmark for the general population in Luxem-bourg and Malta, but the levels of trust are relatively high in these two countries.
Immigrants and descendants of immigrants from North Africa indicate the same levels of trust as ― or higher levels than ― the general population in most countries covered, such as in Belgium, France and Spain. However, they indicate slightly lower levels in the Netherlands. Higher levels can also be observed for respondents with South Asian background in the United Kingdom; no comparisons are available for Cyprus, Greece and Italy. Recent immigrants in Poland and Slovenia indicate considerably higher levels of trust. The EU-MIDIS II findings highlight important links between duration of residence and being part of the first or second generation and levels of trust. This is further discussed below.
Respondents from the Russian minority in Estonia and Lithuania indicate levels of trust that are similar to those expressed by the respective general population. No data permitting such a comparison for Latvia are available.
Low levels of trust are found among Roma respondents. Comparisons with the general population are possible for only four countries: the Czech Republic, Hungary, Portugal and Spain. In these countries, levels of trust in the legal system are similar to, or slightly below, those for the general population. Meanwhile, levels of trust in the police are clearly lower than among the general population in the Czech Republic, Portugal and Spain, but are almost the same as among the general population in Hungary.
101
What do the results show?
Figure 42: Levels of trust in the police, by country and target group (average value on a scale from 0 to 10) a,b,c
109876543210
AT
SSAFR RIMGR
DK FR IT MT SEDE FI IE LU PT UK PL SI
3.6
6.3
7.2
8.2
6.6
5.6
6.47.3 6.3
6.4
6.4 7.6
6.0
6.4
109876543210
TUR NOAFR (S)ASIA
AT SE ITBE NLDE BEDK ES CY ITNL FR EL UK
6.9
5.2
6.6
5.45.0
6.86.3 6.0
7.5
7.3
6.0
7.4
6.25.6
6.1
109876543210
RUSMIN
BG
ROMA
EL HR PT SK EECZ ES RO LT LV
Trust in the police EU-MIDIS II Trust in the police ESS
HU
4.2
5.7
3.7
5.1 5.04.2
5.2
4.1
5.76.1
5.9
4.6
Notes: a EU-MIDIS II (n=25,335); weighted results and European Social Survey (ESS), 2014 (n=34,639). b Question: “Please tell me on a scale of 0–10 how much you personally trust each of the [COUNTRY] institutions
I read out. 0 means you do not trust an institution at all, and 10 means you have complete trust.” c Acronyms for target groups refer to immigrants from [country/region] and their descendants: TUR = Turkey,
SSAFR = Sub-Saharan Africa, NOAFR = North Africa, (S)ASIA = South Asia and Asia, RIMGR = recent immigrants from non-EU countries, RUSMIN = Russian minority, ROMA = Roma minority.
Source: FRA, EU-MIDIS II 2016; European Social Survey, 2014
Second European Union Minorities and Discrimination Survey – Main results
102
Figure 43: Levels of trust in the legal system, by country and target group (average value on a scale from 0 to 10) a,b,c
109876543210
AT
SSAFR RIMGR
DK FR IT MT SEDE FI IE LU PT UK PL SI
4.7
5.9
7.58.0
7.0
4.9
6.3
7.5
6.2 6.2 6.2 6.1
6.06.6
109876543210
TUR NOAFR (S)ASIA
AT SE ITBE NLDE BEDK ES CY ITNL FR EL UK
6.45.5
6.1
4.95.4
6.5 6.6
5.2
7.46.7
5.8
7.36.6
5.8 6.0
109876543210
RUSMIN
BG
ROMA
EL HR PT SK EECZ ES RO LT LV
Trust in the legal system EU-MIDIS II Trust in the legal system ESS
HU
4.1
5.6
3.2
4.3 4.6
3.6
4.7
3.4
5.24.4
4.6
3.3
Notes: a EU-MIDIS II (n=25,256); weighted results and European Social Survey (ESS), 2014 (n=34,622) b Question: “Please tell me on a scale of 0–10 how much you personally trust each of the [COUNTRY] institutions
I read out. 0 means you do not trust an institution at all, and 10 means you have complete trust.” c Acronyms for target groups refer to immigrants from [country/region] and their descendants: TUR = Turkey,
SSAFR = Sub-Saharan Africa, NOAFR = North Africa, (S)ASIA = South Asia and Asia, RIMGR = recent immigrants from non-EU countries, RUSMIN = Russian minority, ROMA = Roma minority.
Source: FRA, EU-MIDIS II 2016; European Social Survey, 2014
103
What do the results show?
Levels of trust are similar among men and women both for EU-MIDIS II respondents and the general population (European Social Survey). Men in the general popula-tion tend to trust the legal system more, which is also observed in EU-MIDIS II. Across the EU-MIDIS II target groups that include immigrants, there are notable gen-erational differences in the levels of trust in the legal system and the police. Second-generation respondents tend to exhibit lower levels of trust than first-genera-tion immigrants. The differences are particularly strong among both target groups covered in France and in the United Kingdom. Descendants of immigrants from Tur-key in Austria are the only second-generation group to indicate higher levels of trust in the legal system than the respective first generation (Figure 44).
EU-MIDIS II findings also show that trust in the legal system is related to having a general interest in politics. However, interest in politics is influenced by several
factors and is strongly gendered. Women tend to be much less interested in politics than men both in the general population and among the target groups cov-ered in EU-MIDIS II. Of all EU-MIDIS II respondents, 43 % of men and 29 % of women indicate being interested in politics (either ‘quite interested’ or ‘very interested’). Interest in politics is lower among Roma respondents (men: 23 %, women: 12 %) and recent immigrants (19 % vs 15 %). Comparatively high rates are observed among male respondents with Turkish background (52 % vs 29 %). Fully understanding the significance of the respective levels of interest would require tak-ing into account country-specific contexts and avail-able opportunities to participate politically. Generally, in almost all countries and target groups, respondents with national citizenship express considerably more interest in politics.60
60 There are some cases with only slightly higher or very similar levels, but the tendency can be observed for almost all countries.
Figure 44: Levels of trust in the legal system, by country, target group and generation a,b,c,d
109876543210
SE -
SSAF
RFR
- SS
AFR
FR -
NOAF
RUK
- SS
AFR
UK -
SASI
ALU
- SS
AFR
BE -
NOAF
RFI
- SS
AFR
NL -
NOAF
RIT
- NO
AFR
DE -
TUR
NL -
TUR
DK -
TUR
DK -
SSAF
RPT
- SS
AFR
SE -
TUR
DE -
SSAF
RBE
- TU
RES
- NO
AFR
AT -
TUR
First generation Second generation
6.6 6.5 6.87.3
7.7
6.6 6.3
8.1
5.74.9
7.5
5.76.6
6.0 6.26.8
7.6
5.8
6.1 6.2
(3.6)4.4 4.7
5.4
6.5
5.4 5.2
7.2
5.0(4.2)
7.1
5.36.2
5.6 5.96.4
7.2
5.7
(6.4)
7.7
Notes: a Out of all groups of immigrants with a considerable share of second-generation immigrants (n=25,256); weighted results.
b Question: “Please tell me on a scale of 0–10 how much you personally trust each of the [COUNTRY] institutions I read out. 0 means you do not trust an institution at all, and 10 means you have complete trust.”
c Acronyms for target groups refer to immigrants from [country/region] and their descendants: TUR = Turkey, SSAFR = Sub-Saharan Africa, NOAFR = North Africa, (S)ASIA = South Asia and Asia, RIMGR = recent immigrants from non-EU countries, RUSMIN = Russian minority, ROMA = Roma minority.
d Results based on a small number of responses are statistically less reliable. Thus, results based on 20 to 49 unweighted observations in a group total or based on cells with fewer than 20 unweighted observations are noted in parentheses. Results based on fewer than 20 unweighted observations in a group total are not published.
Source: FRA, EU-MIDIS II 2016
Second European Union Minorities and Discrimination Survey – Main results
104
2�4�5� Sense of belonging, attachment and social distance
Sense of belonging, attachment and identification
Respondents’ levels of attachment to different regions and entities vary depending on the different circum-stances and situations in which they find themselves. When asked about their level of attachment to their neighbourhood, village/town/city, county/region, coun-try of residence or the European Union, EU-MIDIS II respondents indicated feeling most strongly attached to their country of residence: 77 % feel strongly or very strongly attached to the country in which they live. 16 % are neutral, and only 6 % do not or not at all feel attached. The level of attachment to the EU is lowest; nonetheless, 44 % still feel (very) strongly attached to the EU and 26 % indicate being neutral.
To further explore senses of belonging and identifica-tion, the survey asked respondents to what extent they identify with their country of residence, their country of birth and with Europe.61
In line with results for the question about feelings of attachment, respondents identify less with Europe than with their country of residence. Only 41 % of all respondents feel strongly or very strongly European (with 27 % neutral). By comparison, 67 % identify strongly with their country of residence (and 18 % indicate being neutral). The feeling of identifying with the country of residence is lower among immigrant groups with more recent immigration histories. This is illustrated in Figure 45, which shows the extent to which the various immigrant groups identify with their countries of birth and countries of residence, by gen-eration. For example, first-generation immigrants from (South) Asian countries in Italy, Greece and Cyprus less often indicate feeling Italian, Greek and Cypriot than this group indicates feeling British in the United Kingdom. Due to the recent immigration history of (South) Asians in Italy, Greece and Cyprus, no reliable estimates on the second generation can be made. For the other groups, where comparisons are possible, the second generation always identifies at least as strongly with the country of residence as the first generation, but mostly does so even more strongly.
The extent of identifying with the country of resi-dence varies not only based on generation, but also
61 For example, an immigrant from Turkey residing in Austria was asked: “To what extent do you feel Austrian?” and “To what extent do you feel Turkish?”. Instead of being asked about their own country of birth, the second generation was asked about the country of birth of their mother and/or father. Ethnic minorities were asked to what extent they feel Roma and Russian, respectively, instead of about country of birth.
based on citizenship status, with generation and citi-zenship strongly interlinked. There are no strong dif-ferences between the extents of identifying with the country of residence based on gender, with women indicating that they identify with their country of resi-dence to a slightly higher extent than men.
Among first-generation immigrants, across all target groups and countries, more respondents more strongly identify with their country of birth than with their coun-try of residence. However, identities are not exclusive. In most cases, immigrants identify strongly with both their country of birth and their country of residence (Table 5). Of all first-generation immigrants, one in two (49 %) identify (very) strongly with both countries – even more do so among those with national citizenship (61 %). 15 % identify strongly with their country of birth, and not at all with their country of residence. For 6 %, it is the other way around – they identify with their country of residence, and not at all with their country of birth. Only 1 % indicate not identifying with either of the two countries. Among the second generation, one in two identify (strongly) with their country of residence and with their mother’s and/or father’s country (47 %).
Social distance
Respondents’ level of ‘social distance’ was measured by asking several questions about how comfortable they would feel with people of different groups as their neighbours. The groups asked about include persons with a different religion than the respondents’, persons with another ethnic minority background, persons with-out ethnic minority background, persons with disabili-ties, gay, lesbian or bisexual persons, and transgender or transsexual persons. In addition, respondents were asked about their comfort level with having neighbours from the same ethnic minority as themselves. Not sur-prisingly, almost all respondents felt comfortable with persons of the same ethnic origin (with 98 % indicat-ing they felt comfortable with or neutral about this). Only 2 % felt uncomfortable with neighbours of the same ethnic origin.62
EU-MIDIS II respondents generally indicated being open towards persons belonging to other groups being their neighbours. 95 % to 97 % feel comfort-able with, or neutral about, neighbours with a dif-ferent religion or different ethnic origin, without an ethnic minority background, or with disabilities. The level of acceptance is much lower when it comes to persons with a different sexual orientation. 73 % feel comfortable with, or neutral about, lesbian, gay or
62 For these questions, respondents could indicate their comfort level on a scale from 0 to 10, where 0 means ‘totally uncomfortable’ and 10 means ‘totally comfortable’. Answers ranging from 0 to 4 were categorised as ‘uncomfortable’, 5 as ‘neutral’, and values from 6 to 10 as ‘comfortable’.
105
What do the results show?
bisexual neighbours, and 66 % do so with transgen-der or transsexual persons. However, 4 % and 6 %, respectively, indicated that they did not know if they felt comfortable or refused to answer the question. Ethnic minorities feel uncomfortable with LGBT people as their neighbours at higher rates than other groups of immigrants. Only 47 % of respondents of the Rus-sian minority feel comfortable with, or neutral about, gay, lesbian or bisexual persons as neighbours, and 44 % do so with respect to transgender or transsexual persons. However, a higher level of respondents of the Russian minority also indicated that they did not know
the answer to the question or refused to answer it – at 9 % for both questions. Among Roma respondents, the percentage of those feeling comfortable with or neutral about lesbian, gay or bisexual, or transgender or transsexual, persons as neighbours is lowest in Bul-garia (40 % and 31 %, respectively), Croatia (37 % and 32 %) and Romania (33 % and 28 %). However, many respondents also answered “don’t know” or refused to answer this question. By contrast, Roma in Spain are more open – with 92 % and 89 %, respectively, feeling comfortable with or neutral about lesbian, gay or bisexual, or transgender or transsexual, neighbours.
Figure 45: Identification with country of residence, by country, target group and generation (%) a,b,c,d
FI FI
LU LU
UK UK
DK DK
IT
SE SE
DE DE
IE
PT PT
AT
FR FR
MT
SSAF
R
SE SENL NLDK DKDE DE
AT ATBE BE
TUR
SIPLRI
MGR
CY
UK UK
ELIT
(S)A
SIA
BE BE
NL NLIT IT
ES ESFR FR
NOA
FR
Neutral
First generation Second generation
(Very) stronglyNot (at all)
40 4020 2080 8060 600 0100 100
24 (14)28 (13)48 (73)
2330 4752 4723 (22)25 32
28 34 3811 16 7416 1213 1571 7313 16 (21)72 73
33 (13)30 2837 59
63 16 2130 (13)28 (10)42 77
7 (9)21 73 91
8 (8)17 1875 75
12 24 (13)64 83
28 1328 2443 6326 1629 1845 67
9 626 1765 7730 1730 2440 58
32 1730 3038 53
21 2230 34 36
58
6 14 (7)80 9131 41 28
49 32 1944 24 33
25 2124 2851 5235 (10)27 (4)38 (86)
11 917 1273 7919 (10)24 (32)57 58
12 620 1268 82
(2)
(4)
(6)
Notes: a First generation: n=12,673 and second generation: n=3,365; overall less than 1 % of respondents did not know an answer to the question, these respondents were excluded from the figure above; weighted results.
b Question: “People might see themselves in different ways. The following question is about how you see yourself. On a scale from 1 to 5, where 1 equals ‘not at all’ and 5 ‘very strongly’, to what extent do you feel [SURVEY COUNTRY NATIONAL]?”
c Results based on a small number of responses are statistically less reliable. Thus, results based on 20 to 49 unweighted observations in a group total or based on cells with fewer than 20 unweighted observations are noted in parentheses. Results based on fewer than 20 unweighted observations in a group total are not published.
d Acronyms for target groups refer to immigrants from [country/region] and their descendants: TUR = Turkey, SSAFR = Sub-Saharan Africa, NOAFR = North Africa, (S)ASIA = South Asia and Asia, RIMGR = recent immigrants from non-EU countries.
Source: FRA, EU-MIDIS II 2016
Second European Union Minorities and Discrimination Survey – Main results
106
Notable differences in attitude can be observed among immigrant groups within all target groups across countries of residence, indicating that their views are not directly tied to their respective countries of origin alone. For example, on average, more than every second respondent with Turkish background in Austria would feel uncomfortable with a gay, les-bian or bisexual, or a transgender, neighbour (52 % and 54 %, respectively). Meanwhile, this rate is much lower among respondents with Turkish background in Sweden (16 % and 26 %). The share of those feeling uncomfortable with LGBT neighbours is particularly high among respondents with Sub-Saharan African background in Austria (63 % and 64 %) and Denmark (44 % and 46 %), but much lower in the United King-dom (6 % and 8 %), Finland (13 % and 11 %), and France (14 % and 18 %). This means that countries of destination can also influence immigrants’ views. However, it is also important to remember that the composition of the various target groups varies across destination countries in terms of socio-demographic and other characteristics.
With respect to social distance, on average, sec-ond-generation respondents are slightly more open towards other groups than first-generation respond-ents. Those holding national citizenship of the coun-try in which they reside are on average also slightly more open, particularly towards LGBT people. Women are slightly more open towards LGBT people than
men. The strongest difference in openness can be observed among respondents who indicate having friends with other or no ethnic minority backgrounds; these respondents are more open towards all other categories of people.
Figure 46 presents the percentages of respondents who selected a value between 5 and 10 when asked how comfortable they would feel with people from dif-ferent ethnic groups, or gay, lesbian or bisexual people, as neighbours. It indicates a high level of acceptance towards other ethnic groups. Somewhat lower levels are observed among the Russian minority in Estonia, respondents with Sub-Saharan African background in Austria, and Roma respondents in Slovakia, Greece, and the Czech Republic. As noted above, the nega-tive views towards gay, lesbian and bisexual people outlined in Figure 46 are consistent with the overall lower acceptance levels of gay, lesbian or bisexual people as neighbours among all target groups – with some notable exceptions, such as Roma respond-ents and respondents with North African background in Spain or with Sub-Saharan African background in the United Kingdom.
Some 95 % of all EU-MIDIS II respondents feel com-fortable with or neutral about neighbours who have a different ethnic minority background, and 96 % feel comfortable with or neutral about neighbours who have no ethnic minority background; only 2 %
Table 5: Feeling of attachment to country of residence and country of origin for first- and second-generation respondents (%) a,b
First-generation respondentsTo what extent do you feel [national of country of birth]?Not (at all) Neutral (Very) strongly Total
To what extent do you feel [national of country of residence]
Not (at all) 1 1 15 17Neutral 1 4 16 21(Very) strongly 6 7 49 62Total 8 12 81c 100
Second-generation respondentsTo what extent do you feel [national of country of birth of mother or father]?
Not (at all) Neutral (Very) strongly TotalTo what extent do you feel [national of country of residence]
Not (at all) 2 1 9 12Neutral 2 6 7 15(Very) strongly 11 15 47 73Total 15 22 63 100
Notes: a First generation: n=12,673 and second generation: n =3,365; weighted results. b Question: “People might see themselves in different ways. The following question is about how you see yourself.
On a scale from 1 to 5, where 1 equals ‘not at all’ and 5 ‘very strongly’, to what extent do you feel [SURVEY COUNTRY NATIONAL]?/[national of country of birth]?/[national of country of birth of mother or father]?” The latter was asked for second-generation respondents.
c Number adds up to 81 when decimals are included in the calculation. The table shows rounded numbers.Source: FRA, EU-MIDIS II 2016
107
What do the results show?
tend to feel uncomfortable with the latter. Related to this, most EU-MIDIS II respondents have friends of another or no ethnic minority background (77 % and 82 %, respectively). Figure 47 shows the percentages of respondents who have friends without an ethnic minority background by target group.
The 2014 European Social Survey asked slightly dif-ferent questions regarding friends with different eth-nic backgrounds and views on other ethnic groups
and immigrants. Among the general population, a much larger proportion of respondents said that they did not have any close friends with a different ethnic background. For example, in Sweden, France, Slovenia, the United Kingdom, Belgium, Denmark, Spain, the Netherlands, Austria, Germany, and Fin-land, between 30 % and 59 % of the total population indicate not having any such close friends. In Poland, for example, the majority of the population (79 %) indicated not having any close friends with a different
Figure 46: Respondents’ comfort level with other ethnic minorities and gay, lesbian or bisexual persons as neighbours, by country and target group (%) a,b,c
0
40
20
80
60
10
50
30
100
90
70
ATSSAFR RIMGR
DK FR IT MT SEDE FI IE LU PT UK PL SI
85
35
95
72
94
52
97
85
98
84
98
85
95
61
97
78
94
60
95
65
95
72
99
9295
75
96
73
A person of another ethnic minority background than youA gay, lesbian or bisexual person
0
40
20
80
60
10
50
30
100
90
70
TUR NOAFR (S)ASIAAT SE ITBE NLDE BEDK ES CY ITNL FR EL UK
98
44
99
79
93
64
96
72
98
67
97
82
95
81
99
77
96
90
99
85
97
62
93
46
99
75
94
53
95
80
Second European Union Minorities and Discrimination Survey – Main results
108
Figure 47: Respondents who have friends without an ethnic minority background (%) a,b,c
0
40
20
80
60
10
50
30
100
90
70
AT DE DK FI FR IE
SSAFR TUR NOAFR (S)ASIA RIMGR
IT LU MT PT SE UK AT BE DE DK NL SE BE ES FR IT NL CY EL IT UK PL SI
ROMA RUSMIN
BG CZ EL ES HR HU PT EERO LTSK LV
63
81
77
8886
87
67
84
53
94
74
84
7582
86
70
8884
51
73
8082
88
77
90
59
49
67 6762
74
66
92
66
82 81
51
74 76
9093
Notes: a Out of all respondents (n=25,515); weighted results. b Question: “Do you have friends who do not have a minority background?” c Acronyms for target groups refer to immigrants from [country/region] and their descendants: TUR = Turkey,
SSAFR = Sub-Saharan Africa, NOAFR = North Africa, (S)ASIA = South Asia and Asia, RIMGR = recent immigrants from non-EU countries, RUSMIN = Russian minority, ROMA = Roma minority.
Source: FRA, EU-MIDIS II 2016
Figure 46 (continued)
A person of another ethnic minority background than youA gay, lesbian or bisexual person
0
40
20
80
60
10
50
30
100
90
70
RUSMINBG
ROMAEL HR PT SK EECZ ES HU RO LT LV
87
40
79
58
77
44
9692 93
37
83
50
89
41
84
35
91
33
99
50
69
45
93
56
Notes: a All respondents (n=25,515); weighted results. b Question: “And using a scale from 0 to 10, please tell me how would you feel about having someone from one of
the following groups as your neighbour? 0 means that you would feel ‘totally uncomfortable’ and 10 means that you would feel ‘totally comfortable’.”
c Acronyms for target groups refer to immigrants from [country/region] and their descendants: TUR = Turkey, SSAFR = Sub-Saharan Africa, NOAFR = North Africa, (S)ASIA = South Asia and Asia, RIMGR = recent immigrants from non-EU countries.
Source: FRA, EU-MIDIS II 2016
109
What do the results show?
ethnic background.63 These findings are not surpris-ing. Members of the general population sometimes have fewer opportunities to have contact with people of different backgrounds, because – as is the case in Poland, for example – the share of immigrants or ethnic minorities is very low. However, having friends of different ethnic origins is tied to having more open views on immigrants and their contribution to soci-ety. A look at the general population’s views about whether religious beliefs and practices are under-mined or enriched by immigrants reveals that those with close friends of a different ethnic background have more open views.
Gender equality
Respondents were asked to which extent they agree or disagree with the following four statements:
• Having a job is the best way for a woman to be an independent person
• Both husband and wife should contribute to house-hold income
• Men should take as much responsibility as women for the home and children
• It is important that both girls and boys stay in edu-cation for the same length of time.
80 % of all male respondents agreed with the state-ment that “having a job is the best way for a woman to be an independent person”, while 86 % of female respondents agreed. 85 % of men agreed that both hus-band and wife should contribute to household income, as did 88 % of women. 82 % of men agreed that men should take as much responsibility for the home and children, as did 88 % of women. Finally, most respond-ents agreed with the statement that girls and boys should stay in education for the same time – specifically, 92 % and 91 % of men and women agreed, respectively.
Strong differences can be observed across countries and target groups. Table 6 shows results for the second state-ment (about husband and wife contributing to household income). Agreement with that statement is particularly low among male respondents with South Asian back-ground in Greece (39 %), male respondents with Sub-Saharan African background in Malta (53 %), and male respondents with Turkish background in the Netherlands, where only 59 % agree that both should contribute to
63 FRA’s calculations are based on the European Social Survey wave 7, 2014. These percentages cannot be directly compared to EU-MIDIS II results on the percentages of respondents having friends from other or no ethnic background, because the European Social Survey asked about ‘close friends’, while EU-MIDIS II asked about just ‘friends’.
household income. A comparison with the general popu-lation’s agreement levels with ‘gender equality state-ments’, available from the 2008 European Value Study,64 shows that these are, on average, similar to those of EU-MIDIS II target groups. Groups covered in EU-MIDIS II often show similar or higher levels of agreement – with some exceptions, such as the groups mentioned above. Table 6 provides details.
2�4�6� Effect of discrimination and victimisation on sense of belonging and trust in public institutions
Several factors affect individuals’ feeling of belonging and their levels of trust in institutions. As noted above, most respondents feel strongly attached to and identify strongly with their country of residence – with immigrants also strongly identifying with their country of origin. First-generation respondents’ feelings of attachment and identity are influenced by various factors. The feeling of identification with the country of residence increases over time; is higher among those who hold national citizenship; and is considerably lower among those who have expe-rienced discrimination, harassment and violence. Those who have experienced discrimination on average indicate lower levels of identification with their country of resi-dence than those without such experiences. Meanwhile, discrimination experiences do not affect first- and second-generation respondents’ average levels of identification with their (or their mother’s or father’s) country of origin.
Similarly, respondents with negative experiences regard-ing discrimination, harassment or violence show lower levels of trust in the police and the legal system, as shown in Figure 48.
Figure 49 shows the average levels of trust in the legal system for all EU-MIDIS II countries and target groups. Respondents with discrimination experiences consist-ently – though at varying levels – show lower levels of trust in the legal system among almost all target groups and countries. Among some groups, those with discrim-ination experiences tend not to trust the legal system, while those without such experiences tend to trust the legal system. For example, among Roma respondents in Croatia, those without discrimination experiences indicate an average value of 4.9 (on a scale from 0 to 10, where 10 means ‘complete trust’). By contrast, Roma respondents in Croatia with discrimination experiences in the 12 months preceding the survey show an aver-age value of 3.3.
64 EVS 2008, Integrated Dataset; data downloaded on 26 September 2017.
Second European Union Minorities and Discrimination Survey – Main results
110
Table 6: Agreement to gender equality statement “Both husband and wife should contribute to household income”, by gender, country and target group (%) a,b,c,d
EU-MIDIS II EVS 2008
Men Women Men Women
SSAFR
AT 87 92 81 85
DE 92 95 79 86
DK 93 95 77 80
FI 89 90 82 78
FR 93 93 89 91
IE 95 95 76 77
IT 82 83 85 91
LU 91 94 71 76
MT 53 (38) 90 90
PT 98 99 96 96
SE 95 93 92 93
UK 96 96 72 74
TUR
AT 92 86 81 85
BE 81 87 79 84
DE 79 87 79 86
DK 89 93 77 80
NL 59 65 40 52
SE 92 93 92 93
NOAFR
BE 66 81 79 84
ES 87 86 88 88
FR 89 88 89 91
IT 73 85 85 91
NL 59 68 40 52
(S)ASIA
CY 93 94 89 94
EL 39 n.a. 92 94
IT 71 74 85 91
UK 85 87 72 74
RIMGRPL 70 88 86 86
SI 87 94 88 92
ROMA
BG 89 89 97 97
CZ 74 74 91 91
EL 69 73 92 94
ES 90 90 88 88
HR 85 84 79 86
HU 93 88 92 95
PT 93 90 96 96
RO 73 81 87 88
SK 79 73 94 93
111
What do the results show?
EU-MIDIS II EVS 2008
Men Women Men Women
RUSMIN
EE 89 89 85 89
LT 97 99 86 84
LV 86 89 94 94
Notes: a EU-MIDIS II, men: n=13,261 and women: n=12,253; European Value Study (EVS), men: n=19,160 and women: n=20,669. b Question: “People talk about the changing roles of men and women today. For each of the following statements
I read out, can you tell me how much you agree or disagree with each? Both husband and wife should contribute to household income” It is important to note that in EU-MIDIS II disagreement was offered first and in the EVS agreement. This difference in question design limits the comparability somewhat. Additionally, the remaining respondents in EU-MIDIS II also include those who did not know an answer to this question, which is 2 % of men and 1 % of women. This percentage was not taken into account in the EVS data.
c Results based on a small number of responses are statistically less reliable. Thus, results based on 20 to 49 unweighted observations in a group total or based on cells with fewer than 20 unweighted observations are noted in parentheses. Results based on fewer than 20 unweighted observations in a group total are not published (n.a.).
d Acronyms for target groups refer to immigrants from [country/region] and their descendants: TUR = Turkey, SSAFR = Sub-Saharan Africa, NOAFR = North Africa, (S)ASIA = South Asia and Asia, RIMGR = recent immigrants from non-EU countries, RUSMIN = Russian minority, ROMA = Roma minority.
Source: EU-MIDIS II, 2016; European Value Study (EVS), 2008
Figure 48: Impact of experiences with discrimination, harassment and violence on trust in the legal system (mean value on a scale from 0 to 10)
0
1
2
3
4
5
6
7
8
9
10
Aver
age
leve
l of t
rust
6.4 6.3
5.36.1
4.6
6.6
5.2
6.6
5.3
6.3
4.3
Trust in the legal system Trust in the police
Discrimination Harassment Violence Discrimination Harassment Violence
Not a victimVictim
5.1
Source: FRA, EU-MIDIS II 2016
Table 6 (continued)
Second European Union Minorities and Discrimination Survey – Main results
112
Figure 49: Impact of experiences with discrimination on trust in the legal system, by country and target group (mean value on a scale from 0 to 10)
5.3
6.9
6.1
5.9
3.4
6.7
4.6
7.97.5
6.3
6.7
5.3
6.3
6.9
6.4
3.3
8.3
6.5
6.1
4.9
4.9
7.3
5.0
5.2
5.1
4.7
7.0
4.5
8.3
5.9
5.5
6.4
3.7
6.1
4.9
6.7
7.0
6.1
3.7
7.4
6.9
4.0
6.2
5.2
5.3
2.1
5.0
3.0
6.66.5
5.4
5.7
(4.1)
4.9
6.1
5.0
3.1
7.6
4.9
5.7
3.3
3.8
6.2
4.5
5.1
4.6
5.6
(3.6)
6.2
4.9
5.4
3.5
6.5
4.3
5.5
5.2
(5.7)
2.7
6.6
5.1
0 1 2 3 4 5 6 7 8 9 10Average level of trust
ES − ROMABG − ROMASK − ROMAPT − ROMA
LV − RUSMINLT − RUSMIN
CZ − ROMARO − ROMAHR − ROMAIT − NOAFRIT − SSAFR
HU − ROMAAT − SSAFR
IT − SASIAEE − RUSMIN
NL − TURBE − TUR
NL − NOAFRSI − RIMGRPT − SSAFRFR − SSAFR
BE − NOAFRDK − SSAFREL − ROMAPL − RIMGRES − NOAFRFR − NOAFR
DK − TURCY − ASIA
SE − SSAFRUK − SSAFR
AT − TURSE − TUR
EL − SASIALU − SSAFRIE − SSAFRUK − SASIA
DE − TURDE − SSAFRFI − SSAFR
MT − SSAFR
Not a victimVictim
Source: FRA, EU-MIDIS II 2016
113
What do the results show?
Note for Experts: multivariate analysis on the level of trust in the legal system
The impact of several respondent characteristics on the level of trust was tested in a regression analysis, which calculates the average influence of each selected characteristic on the outcome (level of trust in the legal system on a scale from 0 to 10).
Figure 50 presents the results of the regression model. The dots indicate the size of the estimate of each of the parameters and the horizontal bars indicate the uncer-tainty in the parameter estimates (standard errors). If a dot and its bars are clearly away from the vertical line, this indicates that the parameter has an influence on the level of trust. The baseline indicates the estimated level of trust in the legal system when all parameters take the value of 0, which is a hypothetical situation (not shown).
Focus should be placed on the parameter, where it can be seen that length of residence, gender, citizenship and age do not significantly affect the level of trust across the countries. The results show that, for the level of trust in the legal system, across all target groups, hav-ing experienced any form of discrimination, harassment or violence in the 12 months preceding the survey is related to a lower level of trust. The average level of trust in the legal system – measured on a scale from 0 to 10 – is on average 0.6 points lower among respondents with Turkish background who experienced discrimina-tion, 0.7 points lower for respondents with North Afri-can background, 0.9 points lower for respondents with Sub-Saharan African background, and 1.1 points lower for respondents with (South) Asian background.
Figure 50: Regression analysis on the level of trust in the legal system of first-generation migrants, by target group
SSAFR (n = 4,458) TUR (n = 2,038)
(S)ASIA (n = 1,515) NOAFR (n = 2,571)
−1.5 −1.0 −0.5 0.0 0.5
−1.5 −1.0 −0.5 0.0 0.5
−1.5 −1.0 −0.5 0.0 0.5
−1.5 −1.0 −0.5 0.0 0.5
Age
Citizenship: yes
Experiencedvictimisation
Gender: female
Length of residence
Age
Citizenship: yes
Experiencedvictimisation
Gender: female
Length of residence
Age
Citizenship: yes
Experiencedvictimisation
Gender: female
Length of residence
Age
Citizenship: yes
Experiencedvictimisation
Gender: female
Length of residence
Estimated average increase in baseline of trust (scale from 0 to 10)
Para
met
er
Notes: Out of all first-generation respondents for the four target groups with first- and second-generation respondents. The number of observations is provided in brackets next to the target group. The reference country for (S)ASIA is Cyprus, for NOAFR Belgium, for SSAFR and TUR Germany. Differences for countries in the estimated level of trust are not shown. Horizontal bars indicate 95 % standard errors. The baseline for reference countries is not shown for better readability of the graphs ((S)ASIA = 6.7, NOAFR = 5.9, SSAFR = 7.7, TUR 7.0).
Source: FRA, EU-MIDIS II 2016
Second European Union Minorities and Discrimination Survey – Main results
114
Annex I: EU-MIDIS II methodology
FRA’s EU-MIDIS II survey collected data on immigrants and ethnic minorities’ experiences with and opinions on discrimination, victimisation, social inclusion and integra-tion in all 28 EU Member States.
Target groups of immigrants and descendants of immi-grants (often referred to as first- and second-genera-tion respondents) were identified by asking potential respondents about their country of birth and their par-ents’ country of birth. Clearly defined countries and regions of origin were used for the different groups covered in each of the countries. To be considered a member of one of the target groups of immigrants and descendants of immigrants, respondents either had to be born in one of the selected countries of origin (‘first gen-eration’) or one or both of their parents had to be from one of these countries (‘second generation’). In addition, two selected groups of ethnic minorities are included in selected countries: Roma and the Russian minority.
Groups to be surveyed in each of the countries were selected based on multiple criteria, including the size of the target population, feasibility of carrying out a survey with the respective target population, the group’s risk of experiencing ‘racially’, ‘ethnically’ or ‘religiously’ moti-vated discrimination and victimisation, their vulnerability for being at risk of social exclusion and comparability with previous FRA surveys.
For purposes of the survey, immigrants and descendants of immigrants encompass the following:
• ‘Immigrants’ include persons who were not born in an EU Member State or an EEA/EFTA country (Liech-tenstein, Iceland, Norway and Switzerland), have their usual place of residence in the territory of the EU Member State where the survey was conducted, and had been living in the survey country for at least the previous 12 months.
• ‘Descendants of immigrants’ are persons who were born in one of the current 28 EU Member States or EEA/EFTA countries, whose usual place of residence was in the territory of the EU Member State where the survey was conducted, and who had at least one parent not born in an EU or EEA/EFTA country (Liech-tenstein, Iceland, Norway and Switzerland).
• In some EU Member States, EU-MIDIS II interviewed ‘recent immigrants’, namely, persons who immi-grated to an EU Member State in the 10 years before the survey (i.e. after 2004), whose usual place of residence is in the territory of the EU Member State
where the survey was conducted, and who had been living in the survey country for at least 12 months before the interview. The country of birth of ‘recent immigrants’ can be any country other than the EU-28 and other than the EEA/EFTA countries.
• Ethnic minorities, including Roma and the Russian minority, were included based on self-identification.
EU-MIDIS II covered the following groups under the con-cept ‘immigrants and descendants of immigrants’:
• Immigrants and descendants of immigrants from Turkey (in 6 EU Member States);
• Immigrants and descendants of immigrants from North Africa (in 5 EU Member States);
• Immigrants and descendants of immigrants from Sub-Saharan Africa (in 12 EU Member States);
• Immigrants and descendants of immigrants from South Asia and Asia (in 4 EU Member States);
• Recent immigrants from other non-EU/EFTA coun-tries (in 2 EU Member States);
• Russian minorities (in 3 EU Member States);
• Roma (in 9 EU Member States).
For this report, the results were analysed for persons aged 16 years and older, who self-identified with one of the groups listed above and:
• whose usual place of residence is in the EU Member State surveyed;
• who had been living in private households in the EU Member State surveyed for at least the previous 12 months.65
EU-MIDIS II collected information from 25,515 respond-ents living in 22,690 households. The number of respond-ents ranged from 369 for immigrants and descendants of immigrants from Sub-Saharan Africa in Italy to 1,408 Roma in Romania. The sample sizes were determined based on an optimal allocation with respect to the esti-mated total size of the covered target population in addition to practical considerations. For statistics pro-duced in this report, the samples were weighted by their
65 In a small number of countries, persons who were not living in private households were also included in the sample. For example, in Malta, the target population (immigrants and descendants of immigrants from Sub-Saharan Africa) was very small and, without including persons living in institutional homes, the coverage of this population would have been incomplete.
115
Annex I: EU-MIDIS II methodology
estimated size, which means that country and group comparisons take the estimated total size of the target groups per country into account and do not (directly) reflect the sample sizes.
Ipsos MORI, a large international survey company based in the United Kingdom, undertook the fieldwork for EU-MIDIS II under the supervision of FRA staff, who moni-tored compliance with strict quality control procedures.
The main interview mode for EU-MIDIS II was Computer Assisted Personal Interview (CAPI) – that is, face-to-face interviews administered by interviewers using a computerised questionnaire. The English source ques-tionnaire, developed by FRA, was translated into 22 EU languages as well as into Arabic, Kurdish, Russian, Somali, Tamazight and Turkish.
Interviewers were specially trained for the survey, including cultural and ethical training. Wherever pos-sible or necessary, interviewers with the same ethnic background and/or gender conducted the interviews to increase responsiveness among the target groups.
Coverage and selection of countries of originThe countries of origin for each of the previously described target groups were selected based on con-siderations with respect to their vulnerability of being discriminated against. The detailed list of countries of origin for immigrants and descendants of immigrants used for sampling are listed in the separately published EU-MIDIS II Technical Report. The countries included in EU-MIDIS II per target group cover the majority of immi-grants from these respective groups. The six countries covered in EU-MIDIS II with respect to immigrants from Turkey, host 82 % of all immigrants from Turkey in the EU-28, with most settled in Germany. The countries selected for EU-MIDIS II with respect to immigrants from Sub-Saharan Africa, host roughly 86 % of immigrants from this region. EU-MIDIS II countries cover about 92 % of immigrants from North Africa and about 69 % of immi-grants from South Asia in the EU. Figure 51 provides an overview of the number of immigrants for the selected main groups of immigrants in the countries covered com-pared to other countries in the EU.
Implementation of data collection
Sampling
Most of the target groups in EU-MIDIS II can be consid-ered as ‘hard-to-reach’ for survey research – in terms of being relatively small in size and/or dispersed – and due the absence of sampling frames of the target groups.
Whenever possible, a sample was drawn from a sam-pling frame covering the target population. However, the opportunities to sample the target population differed greatly across Member States due to different availability of sampling frames and distribution of the target group in the countries (i.e. list of persons that can be used to make a controlled representative selection of the target group).
Advanced and new sampling methodologies had to be developed and employed in most countries, and the best possible design was chosen for each target group in each of the countries. For some target groups in some coun-tries, a combination of different methods was used to ensure better coverage of the target population. Detailed descriptions of sampling methods used are published in a dedicated Technical Report.
In general, national coverage in some countries had to be reduced for efficiency reasons. This means that in multi-stage sampling, areas with lower densities of the target population were excluded because screening of the tar-get population would not have been possible. In most countries, areas with densities below a certain threshold had to be excluded. These thresholds vary from areas with fewer than 2.7 % in Cyprus to the exceptional case of 30 % in Estonia. These cut-off points, which were unavoidable due to the need for screening respondents in most countries, limited the overall coverage of the target population in the countries. The median cover-age across countries and target groups was 60 % of the target population.
Weighting
The survey results presented in this report are based on weighted data to reflect the selection probabilities of each household and individual based on the sampling design. The weights also account for the differences in the (estimated) size of the target population in each of the countries.
Where possible, the sample was post-stratified to the regional distribution and population characteristics of the covered target population.66 In Finland and the Neth-erlands, the sample was also adjusted to the gender and age distribution. The sample in the Netherlands was further adjusted according to generation (first- or second-generation), country of origin for immigrants and descendants of immigrants from North Africa, and age.
66 External information and data sources for post-stratification are limited. Therefore, in most countries only region and urbanity were used for post-stratification. For example, in Malta, there is a very low percentage of women among the target group. In the absence of detailed population statistics for the target group in Malta, it is still assumed that women were slightly under-represented in the sample but this cannot be adjusted for with the exception of non-response adjustment.
Second European Union Minorities and Discrimination Survey – Main results
116
Sampling error and confidence intervals
All sample surveys are affected by sampling error, given that the survey interviews only a fraction of the total population. Therefore, all results presented are point estimates underlying statistical variation. Small differ-ences of a few percentage points between groups of respondents have to be interpreted within the range of statistical variation and only more substantial differ-ences between population groups should be considered as actual differences in the total population. Results based on small sample sizes are statistically less reli-able and are flagged in figures and tables (for example, numbers shown in figures are put in brackets) and not interpreted substantially. These include statistics that
are based on samples between 20 and 49 respondents in total. Results based on fewer than 20 respondents are not shown. Results based on cell sizes with fewer than 20 persons are flagged as well.
Figure 52 provides an overview of confidence intervals for selected indicators of EU-MIDIS II. The confidence intervals reflect the uncertainty in the estimates due to sampling and are mainly influenced by the sam-pling design and the sample size. The commonly used 95 % confidence intervals are shown. This means that if the sampling was to be carried out repeatedly, in the long run, 95 % of the intervals of the repeatedly carried out samples would contain the true value in the population.
Figure 51: Number of first-generation immigrants from selected target groups, by countries covered in EU-MIDIS II
Turkey: 82 %
Other
DK
SE
BE
AT
NL
DE
Sub–Saharan Africa: 86 %
OtherMTLUATFI
DKIE
SEDEIT
PTFRUK
North Africa: 92 %
Other
NL
BE
IT
ES
FR
0 1,000,000 2,000,000 2,000,000
South Asia: 69 %
Other
EL
IT
UK
0 1,000,000
0 1,000,000 2,000,000 2,000,0000 1,000,000
Notes: FRA’s calculations are based on data from the 2011 Census (Eurostat Census Hub) and, for Finland and the Netherlands, from Eurostat 2015 (table migr_pop3ctb). All data were extracted in February 2017. The pie charts show the percentage of immigrants hosted in the countries selected for EU-MIDIS II.
Source: FRA, 2017 [based on 2011 Census (Eurostat Census Hub) and Eurostat 2015 (table migr_pop3ctb)]
117
Annex I: EU-MIDIS II methodology
Figure 52: Confidence intervals (95 %) of estimates for selected indicators, by country and target group (%)
12 months 5 years
LT − RUSMINLV − RUSMINEE − RUSMIN
SI − RIMGRUK − SASIABG − ROMAPL − RIMGRUK − SSAFR
CY − ASIAPT − SSAFR
DE − TURSE − TUR
ES − NOAFRBE − TUR
RO − ROMAIT − SASIA
HU − ROMAIT − SSAFR
DK − TURAT − TUR
FR − SSAFRSK − ROMAIE − SSAFR
BE − NOAFRCZ − ROMA
FR − NOAFRIT − NOAFRDE − SSAFRES − ROMAEL − SASIA
HR − ROMAMT − SSAFRSE − SSAFR
NL − TURDK − SSAFRAT − SSAFRFI − SSAFR
PT − ROMANL − NOAFREL − ROMALU − SSAFR
Percentage of respondents experiencing any discriminationbased on ethnic origin in any of the nine areas covered
0 20 40 60 80 0 20 40 60 80
36.2
21.4
25.7
15.1
11.0
11.7
26.6
28.5
15.1
35.2
21.0
4.0
42.8
30.2
14.8
29.7
39.9
26.3
23.8
32.1
17.2
22.1
27.0
14.8
15.8
1.5
44.9
3.4
31.1
41.7
31.6
10.9
40.8
12.7
16.7
31.5
14.0
4.6
25.6
5.0
10.9
45.9
31.3
34.7
23.4
15.6
19.4
34.5
38.0
21.7
46.1
30.4
9.1
52.8
42.4
23.7
38.8
49.2
35.1
32.0
41.6
25.1
37.2
38.7
26.6
30.1
5.5
54.6
8.0
42.6
52.9
43.4
17.7
52.3
20.8
22.1
42.7
22.7
10.1
32.2
11.5
18.2
46.1
34.3
38.3
25.3
19.1
22.3
56.9
46.4
29.0
49.1
31.8
7.8
56.0
54.3
27.2
46.2
54.9
41.3
43.6
44.4
27.9
39.7
44.7
33.3
40.2
3.5
64.4
7.5
41.1
59.8
51.9
19.4
65.2
27.9
26.1
45.3
23.6
15.2
50.3
11.4
18.2
55.9
45.2
48.2
35.3
24.7
32.7
65.2
56.6
36.9
60.0
42.4
14.5
65.7
66.8
36.9
55.9
64.2
50.8
53.2
54.1
36.4
55.7
56.7
47.8
58.6
8.5
73.4
13.8
53.0
70.3
63.5
28.0
76.3
38.0
32.2
56.8
34.0
23.8
57.7
20.1
27.0
Notes: In Luxembourg, a quota sample was carried out and the theory of statistical inference using confidence intervals does not apply. Therefore, while the intervals provide a general indication of uncertainty based on the sample size, the normal interpretation does not apply due to the non-random selection of respondents in Luxembourg.
Source: FRA, EU-MIDIS II 2016
Second European Union Minorities and Discrimination Survey – Main results
118
Annex II: Respondents in the EU-MIDIS II survey
EU-MIDIS II covers a variety of groups, which are very heterogeneous in terms of demographic, socio-economic characteristics and migration history. Table 7 provides an overview of the main characteristics of the target groups by country. The average age of respondents is 40 years, with the Russian minority being the oldest group (on average 51 years) and the group of recent immigrants the youngest group (on average 36 years). Among the group of immigrants and descendants of immigrants from Turkey, the average age of respondents varies from 36 years in Belgium to 40 years in Denmark. The group of immigrants and descendants of immigrants from North Africa is youngest in Spain at 35 years and oldest in France at 40 years. Immigrants and descend-ants of immigrants from South Asia show average ages from 34 in Italy to 40 years in the United Kingdom. The group of first- and second-generation respondents with Sub-Saharan African background is youngest in Malta at 28 years and oldest in the United Kingdom at 42 years on average. Finally, Roma are youngest in Croatia at 35 years and oldest in Portugal at 42 years.
Overall, there is a gender balance in the sample, with 51 % women. The Russian minority shows a larger per-centage of women with 60 %. Yet again, across the tar-get groups and countries, there are strong differences in the share of women and men. With the exception of the Baltic countries, higher shares of women (55 % or
higher) are found among immigrants and descendants of immigrants from Asian countries in Cyprus (64 %), Roma in Spain (56 %) and respondents with Sub-Saha-ran African background in Ireland (56 %) and the United Kingdom (55 %). Several groups represent more men than women, which is related to gendered migration patterns. Lower percentages of women (less than 40 %) are found among first- and second-generation respond-ents with Sub-Saharan Africa background in Denmark (31 %) and Austria (26 %). However, immigrants and descendants of immigrants in Malta (from Sub-Saharan Africa) and Greece (from South Asian countries), in par-ticular, show a very low share of women at 6 % and 5 %, respectively. The latter two groups mainly consist of first-generation immigrants. While – by definition – all recent immigrants are first-generation immigrants, the groups of respondents with South Asian and Sub-Saharan African background include higher shares of first-generation immigrants (three in four). Among the groups of respondents with North African and Turkish backgrounds, 66 % and 62 % are first-generation immi-grants. Lowest percentages of first-generation immi-grants, and therefore highest shares of descendants of immigrants, are found in Belgium and the Netherlands. The share of first-generation immigrants is strongly related to the average length of stay in the countries. Target groups with longer average residence in coun-tries naturally show larger percentages of descendants of immigrants. Similarly, the average length of stay also influences the percentage of those with national citizen-ship and groups with longer residence also have larger shares of citizens on average.
119
Annex II: Respondents in the EU-MIDIS II survey
Table 7: Main characteristics of EU-MIDIS II target groups, by country
Country – target group
Average age (years) Women (%) National
citizen (%)First generation
(%)Average
stay (years)Number of
respondentsCY – ASIA 38 64 18 94 11 436AT – TUR 36 50 63 66 22 578BE – TUR 36 46 82 53 27 628DE – TUR 39 48 40 62 30 919DK – TUR 40 50 61 63 29 400NL – TUR 37 48 89 59 30 617SE – TUR 36 45 84 62 22 402BE – NOAFR 37 45 75 53 24 711ES – NOAFR 35 54 19 95 14 787FR – NOAFR 40 52 63 62 25 846IT – NOAFR 35 41 16 97 14 836NL – NOAFR 38 47 89 61 28 653PL – RIMGR 37 53 19 100 6 429SI – RIMGR 35 46 3 100 6 404BG – ROMA 41 53 n.a. n.a. n.a. 1,078CZ – ROMA 39 51 n.a. n.a. n.a. 817EL – ROMA 36 54 n.a. n.a. n.a. 508ES – ROMA 37 56 n.a. n.a. n.a. 776HR – ROMA 35 52 n.a. n.a. n.a. 538HU – ROMA 39 51 n.a. n.a. n.a. 1,171PT – ROMA 42 52 n.a. n.a. n.a. 553RO – ROMA 40 52 n.a. n.a. n.a. 1,408SK – ROMA 37 49 n.a. n.a. n.a. 1,098EE – RUSMIN 50 66 47 n.a. n.a. 401LT – RUSMIN 54 55 99 n.a. n.a. 404LV – RUSMIN 51 56 57 n.a. n.a. 614EL – SASIA 35 5 0 99 12 515IT – SASIA 34 40 8 99 10 517UK – SASIA 40 50 80 71 21 668AT – SSAFR 32 26 10 97 7 476DE – SSAFR 40 48 51 84 20 500DK – SSAFR 34 31 58 87 17 451FI – SSAFR 33 42 59 83 13 502FR – SSAFR 38 51 59 75 18 794IE – SSAFR 36 56 72 97 12 425IT – SSAFR 34 40 17 94 12 369LU – SSAFR 34 49 18 78 10 402MT – SSAFR 28 6 2 100 5 411PT – SSAFR 42 48 43 87 23 525SE – SSAFR 33 42 56 88 12 400UK – SSAFR 42 55 78 67 23 548Total 40 51 57 68 23 25,515
Note: n.a. = not applicable.Source: FRA, EU-MIDIS II 2016
Second European Union Minorities and Discrimination Survey – Main results
120
Table 8: Most important countries of origin of first-generation immigrants per country and target group
Country – target group Country/region of birth Number of respondents
% within country and target group
AT - SSAFR Nigeria 275 60
AT - SSAFR Other 127 28
AT - SSAFR Ghana 34 7
AT - SSAFR Kenya 24 5
DE - SSAFR Other 107 25
DE - SSAFR Eritrea 81 19
DE - SSAFR Ghana 78 18
DE - SSAFR Togo 46 11
DE - SSAFR Ethiopia 41 9
DE - SSAFR Nigeria 30 7
DE - SSAFR Cameroon 26 6
DE - SSAFR Senegal 25 6
DK - SSAFR Somalia 354 91
DK - SSAFR Other 37 9
FI - SSAFR Somalia 139 33
FI - SSAFR Other 113 27
FI - SSAFR Nigeria 47 11
FI - SSAFR Ghana 39 9
FI - SSAFR Ethiopia 33 8
FI - SSAFR Cameroon 30 7
FI - SSAFR Kenya 20 5
FR - SSAFR Other 173 28
FR - SSAFR Senegal 66 11
FR - SSAFR Congo 63 10
FR - SSAFR Cote d’Ivoire 60 10
FR - SSAFR Mali 44 7
FR - SSAFR Guadeloupe 40 7
FR - SSAFR Comoros 33 5
FR - SSAFR Martinique 30 5
FR - SSAFR Democratic Republic of the Congo 28 5
FR - SSAFR Cameroon 25 4
FR - SSAFR Haiti 25 4
FR - SSAFR Cabo Verde 22 4
IE - SSAFR Nigeria 218 52
IE - SSAFR Other 93 22
IE - SSAFR Somalia 42 10
IE - SSAFR Democratic Republic of the Congo 23 6
IE - SSAFR Angola 21 5
IE - SSAFR Congo 20 5
IT - SSAFR Senegal 122 34
IT - SSAFR Other 113 32
121
Annex II: Respondents in the EU-MIDIS II survey
Country – target group Country/region of birth Number of respondents
% within country and target group
IT - SSAFR Nigeria 71 20
IT - SSAFR Ghana 48 14
LU - SSAFR Cabo Verde 126 40
LU - SSAFR Other 114 37
LU - SSAFR Cameroon 25 8
LU - SSAFR Senegal 24 8
LU - SSAFR Guinea-Bissau 23 7
MT - SSAFR Somalia 298 73
MT - SSAFR Other 61 15
MT - SSAFR Eritrea 32 8
MT - SSAFR South Sudan 20 5
PT - SSAFR Cabo Verde 231 50
PT - SSAFR Guinea-Bissau 106 23
PT - SSAFR Angola 68 15
PT - SSAFR Sao Tome and Principe 52 11
PT - SSAFR Other 9 2
SE - SSAFR Somalia 133 37
SE - SSAFR Other 126 35
SE - SSAFR Eritrea 43 12
SE - SSAFR Ethiopia 30 8
SE - SSAFR Nigeria 26 7
UK - SSAFR Other 172 43
UK - SSAFR Jamaica 93 23
UK - SSAFR Somalia 50 12
UK - SSAFR Nigeria 44 11
UK - SSAFR Ghana 42 10
AT - TUR Turkey 410 100
BE - TUR Turkey 332 100
DE - TUR Turkey 597 100
DK - TUR Turkey 260 100
NL - TUR Turkey 263 100
SE - TUR Turkey 265 100
BE - NOAFR Morocco 406 89
BE - NOAFR Other 26 6
BE - NOAFR Algeria 24 5
ES - NOAFR Morocco 731 97
ES - NOAFR Algeria 22 3
FR - NOAFR Algeria 240 45
FR - NOAFR Morocco 219 41
FR - NOAFR Tunisia 72 13
FR - NOAFR Other 5 1
Table 8 (continued)
Second European Union Minorities and Discrimination Survey – Main results
122
Country – target group Country/region of birth Number of respondents
% within country and target group
IT - NOAFR Morocco 544 67
IT - NOAFR Tunisia 119 15
IT - NOAFR Egypt 104 13
IT - NOAFR Algeria 34 4
IT - NOAFR Other 14 2
NL - NOAFR Morocco 275 93
NL - NOAFR Other 20 7
CY - ASIA Philippines 108 26
CY - ASIA Other 79 19
CY - ASIA Vietnam 71 17
CY - ASIA Syrian Arab Republic 64 15
CY - ASIA Georgia 45 11
CY - ASIA Lebanon 29 7
CY - ASIA China 24 6
EL - SASIA Pakistan 302 59
EL - SASIA Bangladesh 167 33
EL - SASIA India 42 8
EL - SASIA Other 1 0
IT - SASIA Bangladesh 212 42
IT - SASIA India 141 28
IT - SASIA Pakistan 124 24
IT - SASIA Sri Lanka 24 5
IT - SASIA Other 8 2
UK - SASIA Pakistan 274 58
UK - SASIA Bangladesh 147 31
UK - SASIA Sri Lanka 44 9
UK - SASIA Other 11 2
PL - RIMGR Ukraine 215 50
PL - RIMGR Other 87 20
PL - RIMGR Belarus 41 10
PL - RIMGR Vietnam 36 8
PL - RIMGR Russian Federation 30 7
PL - RIMGR Turkey 20 5
SI - RIMGR Bosnia and Herzegovina 193 48
SI - RIMGR The former Yugoslav Rep. of Macedonia 78 19
SI - RIMGR Serbia 45 11
SI - RIMGR Kosovo 44 11
SI - RIMGR Other 44 11
Note: All countries of birth with fewer than 20 respondents per country were included in the category ‘Other’.Source: FRA, EU-MIDIS II 2016
Table 8 (continued)
123
Annex III: Awareness of equality bodies in the EU-MIDIS II surveyTable 9: Awareness of equality bodies in each EU Member State (%) a,b
EU Member State “Have you ever heard of the [NAME OF EQUALITY BODY]?” Yes No Don’t know
Austria1 Gleichbehandlungskommission 9 90 (0)2 Gleichbehandlungsanwaltschaft 9 91 (0)3 Verein ZARA, Zivilcourage und Anti-Rassismus-Arbeit 15 85 (0)
Belgium
1 Le Centre interfédéral pour l’égalité des chances et la lutte contre le racisme et les discriminations/Le Centre fédéral migrationInterfederaal centrum voor gelijke kansen en bestrijding van discriminatie en racisme/Federaal Migratiecentrum
32 66 (2)
2 Gelijke Kansen in Vlaanderenc 26 74 -3 Institut pour l’égalité des femmes et des hommes
Instituut voor de gelijkheid van vrouwen en mannen 36 63 (1)
Bulgaria1 Комисия за защита от дискриминация 21 77 (1)2 Омбудсман на Република България 27 68 (1)
Croatia1 Pravobranitelj za osobe s invaliditetom 37 62 (2)2 Pravobraniteljica za ravnopravnost spolova 33 65 (2)3 Pučki pravobranitelj 33 66 (1)
Cyprus
1 Αρχή Ισότητας 49 31 192 Επιτροπή για την Ισότητα των Φύλων στην
Απασχόληση και στην Επαγγελματική Εκπαίδευση 40 38 21
3 Γραφείο Επιτρόπου Διοικήσεως (Ombudsman) 27 50 21Czech Republic 1 Veřejný ochránce práv 52 46 (1)
Denmark1 Ligebehandlingsnævnet 31 67 (1)2 Institut for Menneskerettigheder 60 38 (1)
Estonia1 Soolise võrdõiguslikkuse ja võrdse kohtlemise volinik
Уполномоченный по гендерному равноправию и равному обращению
24 69 5
Finland1 Yhdenvertaisuusvaltuutettu 21 77 (1)2 Tasa-arvovaltuutettu 38 59 (1)
France 1 Le défenseur des droits antérieurement “La Halde” 36 64 (0)
Germanyd
1 Antidiskriminierungsstelle des Bundes 26 73 12 Landesstelle für Gleichbehandlung -
gegen Diskriminierung – Berlin 18 81 (1)
3 Amt für multikulturelle Angelegenheiten (AmkA) – Frankfurt am Main 17 83 (1)
4 Antidiskriminierungsstelle für Menschen mit Migrationshintergrund (AMIGRA) – München 12 87 (1)
Greece1 Συνήγορο του Πολίτη 25 70 42 Επιθεώρηση εργασίας 33 61 63 Επιτροπή Ίσης Μεταχείρισης 6 84 7
Hungary1 Egyenlő Bánásmód Hatóság 28 71 (1)2 Alapvető Jogok Biztosának Hivatala 17 82 (1)
Ireland1 Equality Tribunal 37 61 (1)2 Irish Human Rights and Equality Commission 40 57 (2)3 Office of the Ombudsman 45 55 -
Second European Union Minorities and Discrimination Survey – Main results
124
EU Member State “Have you ever heard of the [NAME OF EQUALITY BODY]?” Yes No Don’t know
Italy1 Ufficio Nazionale Antidiscriminazioni Razziali (UNAR) 8 89 22 Consigliera Nazionale di Parità 10 88 2
Latvia 1 Latvijas Republikas Tiesībsarga birojsБюрo омбудсмена Латвийской Республики 56 37 7
Lithuania 1 Lygių galimybių kontrolieriaus tarnybaКонтроллер равных возможностей 52 45 (2)
Luxembourg 1 Centre for Equal TreatmentCentre pour l’Égalité de Traitement 12 87 (0)
Malta1 National Commission for the Promotion of Equality (NCPE) 6 90 (3)2 National Commission Persons with Disability (KNPD) 6 92 (3)
Netherlands1 College voor de Rechten van de Mens 30 70 (0)2 Antidiscriminatiebureau 36 63 (0)
Poland 1 Rzecznik Praw Obywatelskich 52 44 (1)
Portugal
1 Comissão para a Cidadania e a Igualdade de Género 25 74 (1)2 Comissão para a Igualdade no Trabalho e no Emprego 26 72 (2)3 Alto Comissariado para as Migrações/Comissão para
a Igualdade e Contra a Discriminação Racial 22 77 (1)
Romania 1 Consiliul National pentru Combaterea Discriminarii 23 75 (2)Slovakia 1 Slovenské národné stredisko pre ľudské práva 27 69 4Slovenia 1 Zagovorniku načela enakosti 10 89 0
Spain
1 Consejo para la Eliminación de la Discriminación Racial o Étnica 3 94 (2)
2 Servicio de Asistencia a Víctimas de Discriminación Racial o Étnica 5 92 3
Sweden1 Diskrimineringsombudsmannen 42 56 (2)2 Sveriges antidiskrimineringsbyråer 22 75 (3)
UK1 Equality and Human Rights Commission 53 46 (1)2 Equality Commission for Northern Ireland 8 91 (1)3 The Equal Rights Trust 25 74 (1)
Notes: a Out of all respondents (n=25,440); weighted results. Due to included but not shown answers, such as “does not understand the question” and “don’t know”, the percentage does not always add up to 100 %.
b Results based on a small number of responses are statistically less reliable. Thus, results based on 20 to 49 unweighted observations in a group total or based on cells with fewer than 20 unweighted observations are noted in parentheses. Results based on fewer than 20 unweighted observations in a group total are not published.
c The “Gelijke Kansen in Vlaanderen” was only asked to those being interviewed in Dutch. d In Germany, all respondents were asked about two equality bodies – Equality Body 1 and then either body 2, 3 or
4 depending on where they were interviewed in the country.Source: FRA, EU-MIDIS II 2016
125
ReferencesCouncil Conclusions (2003/C 134/02) of 5 May 2003 on reference levels of European average performance in education and training (Benchmarks), OJ C 134, 7 June 2003.
Council Directive 2000/43/EC of 29 June 2000 imple-menting the principle of equal treatment between persons irrespective of racial or ethnic origin, OJ L 180, 19 July 2000.
Council Directive 2003/86/EC of 22 September 2003 on the right to family reunification, OJ L 251, 3 October 2003.
Council Directive 2003/109/EC of 25 November 2003 concerning the status of third-country nationals who are long-term residents, OJ L 16, 23 January 2004.
Council Framework Decision 2008/913/JHA of 28 Novem-ber 2008 on combating certain forms and expressions of racism and xenophobia by means of criminal law, OJ L 328, 6 December 2008.
Council of Europe (2012), Descriptive Glossary of terms relating to Roma issues, Strasbourg, Council of Europe.
Council of the European Union, (2011), An EU Framework for National Roma Integration Strategies up to 2020 – Council Conclusions, OJ C 258, 2 September 2011.
Council of the European Union (2013), Council Recom-mendation on effective Roma integration measures in the Member States, OJ C 378, 24 December 2013.
Council of the European Union, Justice and Home Affairs (2004), Press Release 2618th Council Meeting, C/04/321, 14615/04 (Presse 321).
Directive 2012/29/EU of the European Parliament and of the Council of 25 October 2012 establishing minimum standards on the rights, support and protection of vic-tims of crime, and replacing Council Framework Decision 2001/220/JHA, OJ L 315, 14 November 2012.
European Commission (2016), Communication from the Commission to the European Parliament, the Council, the European Economic and Social Com-mittee and the Committee of the Regions: Action Plan on the Integration of Third-Country Nationals, COM (2016) 377 final.
European Commission (2017), Countering online hate speech – Commission initiative with social media platforms and civil society shows progress, Brussels, 1 June 2017.
Eurostat (2011), Indicators of immigrant integration: A pilot study, Luxembourg, Publications Office.
Eurostat (2017), Migrant integration: 2017 edition, Luxembourg, Publications Office of the European Union (Publications Office).
European Union Agency for Fundamental Rights (FRA) (2010), EU-MIDIS, European Union minori-ties and discrimination survey: Main results report, Luxembourg, Publications Office.
FRA (2014), Violence against women: an EU-wide sur-vey. Main results, Luxembourg, Publications Office.
FRA (2016), Second European Union Minorities and Discrimination Survey: Roma - Selected findings, Luxembourg, Publications Office.
FRA (2017a), Fundamental Rights Report 2017, Luxembourg, Publications Office.
FRA (2017b), Second European Union Minorities and Discrimination Survey: Muslims – Selected findings. EU-MIDIS II, Luxembourg, Publications Office.
Huddleston, T., Niessen, J. and Dag Tjaden, J. (2013), Using EU Indicators of Immigrant Integration, Final Report for Directorate-General for Home Affairs.
OECD and European Union (2015), Indicators of Immigrant Integration 2015 – Settling In, Paris, OECD Publishing.
HOW TO OBTAIN EU PUBLICATIONS
Free publications:• one copy:
via EU Bookshop (http://bookshop.europa.eu);
• more than one copy or posters/maps: from the European Union’s representations (http://ec.europa.eu/represent_en.htm); from the delegations in non-EU countries (http://eeas.europa.eu/delegations/index_en.htm); by contacting the Europe Direct service (http://europa.eu/europedirect/index_en.htm) or calling 00 800 6 7 8 9 10 11 (freephone number from anywhere in the EU) (*).
(*) The information given is free, as are most calls (though some operators, phone boxes or hotels may charge you).
Priced publications:• via EU Bookshop (http://bookshop.europa.eu).
FRA’s online data explorer tool allows for quick access to the full survey data, see http://fra.europa.eu/en/publications-and-resources/data-and-maps/.
Based on the EU-MIDIS II survey results, two Selected findings reports have been published – one focusing on Roma and one on Muslims� In addition, a detailed Technical report is available�
Second European Union Minorities and Discrim
ination Survey – Muslim
s – Selected findingsFRA
EU-MIDIS II
Second European Union Minorities and Discrimination Survey
Muslims – Selected findings
Second European Union Minorities and Discrim
ination Survey – Technical reportFRA
EU-MIDIS II
Second European Union Minorities and Discrimination Survey
Technical report
Second European Union Minorities and Discrim
ination Survey | Roma – Selected findings
FRA
EU-MIDIS II
Second European Union Minorities and Discrimination Survey
Roma – Selected findings
Seventeen years after adoption of EU laws that forbid discrimination, immigrants, descendants of immigrants, and minority ethnic groups continue to face widespread discrimination across the EU and in all areas of life – most often when seeking employment. For many, discrimination is a recurring experience. Hate-motivated harassment too remains a scourge. While individuals believe their ethnic or immigrant background is the main reason for facing discrimination, they identify their names, skin colour and religion as additional triggers. Not surprisingly, experiences with discrimination and hate-motivated harassment and violence chip away at individuals’ trust in public institutions and undermine feelings of attachment to their country of residence.
These are just some of the findings of FRA’s second European Union Minorities and Discrimination Survey (EU-MIDIS II), which collected information from over 25,500 respondents with different ethnic minority and immigrant backgrounds across all 28 EU Member States. It follows up and expands on FRA’s first major EU-wide survey on minorities’ and migrants’ experi-ences, conducted in 2008. The survey focuses on discrimination in different settings, police stops, criminal victimisation, rights awareness and societal participation.
HELPING TO MAKE FUNDAMENTAL RIGHTS A REALITY FOR EVERYONE IN THE EUROPEAN UNION
FRA - EUROPEAN UNION AGENCY FOR FUNDAMENTAL RIGHTSSchwarzenbergplatz 11 – 1040 Vienna – AustriaTel. +43 1580 30-0 – Fax +43 1580 30-699fra.europa.eu – info@fra.europa.eufacebook.com/fundamentalrightslinkedin.com/company/eu-fundamental-rights-agencytwitter.com/EURightsAgency ISBN 978-92-9491-762-1
doi:10.2811/268615 TK-AN
-17-003-EN-N
top related